information

I read Chris Anderson's book because it was, well, "Free". The book's thesis is simple: Sometimes people profit by giving things away.

I have been, for several years now, making scientific knowledge available for no cost to any readers who care to come by my site. In academic circles, this practice is ordinarily considered to be insane. Therefore, whenever I come across anything explaining why blogging isn't such a stupid idea, I put it right into my files. That's for Luddites on future promotion committees.

How do I review the book without making points like a Slashdot comment thread?

Somewhere in the book, Anderson wrote his plan for making money from Free: Get businesses to pay for the Chris Anderson "Free" seminar. The short business profiles and catchy anecdotes in the book were pretty well crafted as advertisements for the seminar. But beneath the chrome, there are some interesting -- sometimes wacky -- ideas about the nature of human economic interactions.

Software publisher O'Reilly is running an interview with David Dooling, data chief of the Genome Sequencing Center at Washington University: "Sequencing a genome a week". If you want a little background on the current challenges in genomics, the history of genome sequencing technologies, and the infrastructure that allows modern bioinformatics, it's a really nice interview. Dooling is a speaker at the upcoming Open Source Convention (OSCON).

A sample:

James Turner: It sounds like there are a lot of informatics challenges with genomic data. There's the computational challenge of doing the sequence, which you mentioned. There's a challenge of managing the resulting data and finding meaning in it. And then there's the challenge of applying that understanding to a larger population. First of all, did I miss any of the challenges? And second of all, what are the unique problems in each set of those?

David Dooling: Well, let me talk a little bit about each of the ones you did mention, and maybe that'll bring up some that you didn't. So as far as just analyzing the data and generating the data, that is computationally intensive because essentially what you're getting off of these new instruments is pictures, images. And you need to apply algorithms to detect features in those images and then translate those features accounting for different vagaries of chemistry, and then resulting in a sequence, a series of basic Gs, As, Cs and Ts, the building blocks of DNA. Once you have that information, there's a whole host of secondary analysis, or analysis of biological relevance if you want to think of it that way, that need to happen, and those are project-specific. So for some sorts of projects, for example a cancer project, you would want to find all of the ways that the DNA that you sequenced differs from the reference and then take -- for the tumor, let's say. And then for the normal, do the same thing. And then for all of those variants, find out which ones are unique to the tumor genome as compared to the normal genome.

It's not an interview about biology; it's about technology and how people are working to enable us to test more and more detailed biological questions.

I'm a big booster of the idea that human demographic expansion helped drive our recent evolution. So you might expect me to like the new paper by Adam Powell, Stephen Shennan and Mark Thomas, titled, "Late Pleistocene demography and the appearance of modern human behavior." Yet, I see a lot of weaknesses in the paper. I think the paper tries to sidestep several issues about "modern human behavior" that ought to be tackled head-on. In the end, the model in the paper can't describe the data the authors want to consider. Maybe they should have adopted a different model; maybe different data.

I've taken a lot of notes about this -- too many for me to share, but I wanted to review the basic exposition of the paper, including why the authors think demography may determine technological change during the Late Pleistocene. I might post other notes later on the issue of genetic modeling of demography and its relevance for archaeology.

I was talking with a scientist last week who is in charge of a massive dataset. He told me he had heard complaints from many of his biologist friends that today's students are trained to be computer scientists, not biologists. Why, he asked, would we want to do that when the amount of data we handle is so trivial?

Now, you have to understand, to this person a dataset of 1000 whole genomes is trivial. He said, don't these students understand that in a few years all the software they wrote to handle these data will be obsolete? They certainly aren't solving interesting problems in computer science, and in a short time, they won't be able to solve interesting problems in biology.

I said, well, yeah. I've been through this once already -- fifteen years ago, the hot thing was setting up a wet lab for sequencing -- or worse, RFLP. That sure looked like a lot of data at the time, and a lot of students spent a lot of time figuring out how to do it. Some of them successfully started careers, got grants, and moved on with the times. Others fell by the wayside. Meanwhile, clusters of people at the DOE, Whitehead Institute, Wellcome Trust and several private companies were spending their time figuring out faster and faster ways of automating sequencing. Now one machine can do the work of ten thousand 1990's graduate students.

Anyway, I've was thinking about that conversation. And then I ran across an article by Nova Spivack, describing the new Wolfram Alpha.

Stephen Wolfram is building something new -- and it is really impressive and significant. In fact it may be as important for the Web (and the world) as Google, but for a different purpose. It's not a "Google killer" -- it does something different. It's an "answer engine" rather than a search engine.

...

Wolfram Alpha is a system for computing the answers to questions. To accomplish this it uses built-in models of fields of knowledge, complete with data and algorithms, that represent real-world knowledge.

For example, it contains formal models of much of what we know about science -- massive amounts of data about various physical laws and properties, as well as data about the physical world.

Based on this you can ask it scientific questions and it can compute the answers for you. Even if it has not been programmed explicity to answer each question you might ask it.

This sounds very pie-in-the-sky. And indeed, commenters on the article (as well as this article by Cycorp head Doug Lenat) come up with lots of questions that would be impossible for such a system to answer.

But I'm not really interested in the things that will stump the system. Compared to restaurant reviews and kinship systems, bioinformatics is pretty simple. Right now, there are two things that make it a multi-year effort to learn: mutually incompatible databases, and the various kludges necessary to model ascertainment bias.

I'm a Mathematica user, and am familiar with its theorem-proving capabilities. Mathematica already has genome lookup utilities, which I use quite often -- it's just easier to do a lookup on my own system than to plow through two or three webpages to get to the query. It really wouldn't take that much to bring intelligent and interactive genome analysis into the system.

Alpha could turn into an online robot armed with basic genetics knowledge. And if not Alpha -- genetics is a logical priority for Wolfram, but it may not be the first or primary one -- certainly some other system using similar technology will emerge. Put it to work on public databases of genetic information, and you have a system that can resolve the incompatibilities by adding semantic knowledge. A bit of effort on existing databases would allow the resolution of discrepancies in ascertainment. Or, more likely, another couple of years of whole-genome sequencing will solve most of ascertainment biases by drowning them in new data.

So it's not a stretch for me to imagine a year from now entering this search query:

"List all human genes with significant evidence of positive selection since the human-chimpanzee common ancestor, where either the GO category or OMIM entry includes 'muscle'"

It seems to me that bioinformatics is what generates the output to that query. What you do with the output of that query is evolutionary biology.

So that raises the obvious question. Tomorrow's high-throughput plain-English bioinformatics tool will do the work of ten thousand 2009 graduate students. If a freely-available (or heck, even a paid) service can do the bioinformatics, what should today's graduate students be learning?

UPDATE (2009-03-19):

Some folks have interesting reactions to this post, including Thomas Mailund and Dan MacArthur. They make good points.

I will add that I'm not arguing against modeling or simulation in biology. There are lots of interesting things in evolutionary biology you can do -- must do, in all practical terms -- with computers. But I don't like the five-year degree program in genetics where only one semester is given to population genetics, and most of the student's time is spent learning scripting, doing data entry, and figuring out ten or twelve database formats.

I come back to my first example -- fifteen years ago, people were telling you how essential and wonderful sequencing would always be. If you're pursuing a five-year degree program and two or three years of postdoc, I hope you're thinking about what skills you'll need fifteen years from now.

If you're looking for a way to waste your time today, you might check out The Economist's online debate, which focuses on the question of whether the world is getting more or less cultured. Or as they put it, "smarting up or dumbing down":

Intelligent Life, The Economist's quarterly sister magazine, has been looking into what is happening to culture in Britain. The editor, Tim de Lisle, presents a mass of evidence that makes a seemingly irrefutable case: all over the country, more people are going to museums, visiting literary festivals and listening to classical music than ever before. If that isn't wising up, it is hard to know what is.

Susan Jacoby, a scholar whose career began as a reporter on the Washington Post and whose writing now focuses on American intellectual history, sees no reason for Westerners to pat themselves on the back. The education bar, in the Anglo-American world, at least, she says, is being set lower and lower. Fewer and fewer people read books; instead they just hoover up information on the internet. After she wrote an article for her former paper on the decline of reading, she received a deluge of emails from people who said they were proud that they never read books at all. They couldn't see the point.

The recurring issue in the debate seems to be whether people are using information in a deeper or more superficial way. Since both these terms are laden with moral value (always better to be deep than superficial, right?), one may wonder whether the real question isn't whether we feel better about ourselves or not.

Indeed, the two participants devolve immediately into schoolmarmy arguments about whether "high culture" is thriving or not. So we have "increasing attendance at museums" on one side the balance and "decreasing market for hardbound fiction" on the other. Blah.

It would be more interesting to consider the biocultural question: If our culture presents us with more information, do we actually get better at using it over time? There's no mention of the Flynn Effect in the debate, but it seems very relevant -- especially considering the worry that the Brits are "dumbing down".

Zenobia Jacobs and colleagues have a paper in this week's Science that provides age estimates for two of the MSA industries of Southern Africa: the Howieson's Poort and Still Bay industries. Here's the abstract:

The expansion of modern human populations in Africa 80,000 to 60,000 years ago and their initial exodus out of Africa have been tentatively linked to two phases of technological and behavioral innovation within the Middle Stone Age of southern Africa—the Still Bay and Howieson's Poort industries—that are associated with early evidence for symbols and personal ornaments. Establishing the correct sequence of events, however, has been hampered by inadequate chronologies. We report ages for nine sites from varied climatic and ecological zones across southern Africa that show that both industries were short-lived (5000 years or less), separated by about 7000 years, and coeval with genetic estimates of population expansion and exit times. Comparison with climatic records shows that these bursts of innovative behavior cannot be explained by environmental factors alone.

It's a dating paper, and I like the dating parts. The review of why these two MSA industries are important, I think, overstates the issues to a considerable extent. Yes, there are some interesting elements of the two industries, but these are paralleled in some other MSA industries, both earlier and later, in East and North Africa -- not to mention the Neandertal-associated Middle Paleolithic industries of the Near East and Europe. There is no reason at all to suppose that Howieson's Poort (or the earlier Still Bay) was made by people who embarked from southern Africa on an "out of Africa exodus." The southern African sites are important enough for what they tell us about cultural variability; I don't see the need to exaggerate their significance to the global story.

In many ways, the paper relies on similar methods as found in the 2007 paper by Michael Waters and Thomas Stafford, "Redefining the age of Clovis." In that paper, the authors applied a statistical model to new and existing radiocarbon dates, which allowed them to conclude that the age interval represented by Clovis sites is relatively narrow -- probably as little as 200 years.

That conclusion has not gone unchallenged (e.g., Haynes et al. 2007), in particular on the basis of some earlier dates which might indicate an initially rare Clovis lasted for some time before a brief florescence. Anytime we have to deal with dates from different methods or different laboratories, there is the potential that some will be systematically different. Should we dismiss outliers? Or are they essential evidence of a more extensive time range, during which an industry was relatively rare? Hamilton and Buchanan (2007) found a spatial gradient in Clovis radiocarbon dates, suggesting that they represented a wave of advance from north to south. That observation doesn't refute the short chronology, it refines our notion of how long an industry should persist, and shows that it need not represent a spatially uniform population.

In the current paper on Howieson's Poort and Still Bay dating, Jacobs and colleagues took the approach of systematically providing new OSL dates for nine sites. That deals with ambiguity about earlier dates and different methods quite simply: The authors did not rely on dates from other labs and sources. They do present a figure that puts other labs' dates in the context of their own results (they are consistent with the paper's conclusions), but these do not form the main interpretive context.

The essential picture from the paper is figure 4:

This shows the cluster of dates that fit into Howieson's Poort phase, all consistent with a range from around 60,000 to 65,000 years ago, a cluster for the initial post-Howieson's Poort deposits, most consistent with a date around 57,000 years ago, and a smaller cluster of earlier, Still Bay levels. Considering the problems that have plagued OSL dating up to now, this is an impressive level of consistency. Comparing many dates from different sites gives a solid impression of a short time span for the technology.

Unlike the case of Clovis, Jacobs and colleagues found no spatial pattern in the dates, even though they did look. The figure also shows paleoclimate evidence from ice cores; the Howieson's Poort appears to correspond to a long warming period, but it spans the range of climate from cold to warm. That's what the abstract means when it says that environmental factors do not suffice to explain the industry.

I think the dates are important because of what they can tell us about cultural and biological variability within the MSA. From genetics, we know that the MSA African population was apparently structured, with a clear possibility that the genetic differentiation was once higher than today. If so, we might expect long-lasting cultural differences between African regions. We will need better dates across Africa---not just southern Africa---to really compare regions with each other. Howieson's Poort and Still Bay cultures are a start in this process.

The short duration of the two industries is a very important fact. It was already suspected that the two existed for only a short time -- they are not found in every well-stratified site, and their recognition depends on a few relatively rare artifacts. A rare, high-information artifact is useful as a type fossil, but it is not likely to have persisted for very long in the cultural history of an ancient people.

The data seem to indicate that Howieson's Poort lasted around 5000 years, and spanned an area of between 1.5 and 2 million square kilometers. That falls well within the ranges of time span and duration for the industries of the European Upper Paleolithic, and for that matter the later Middle Paleolithic of Europe. The Still Bay, even shorter and smaller, is also within this range. It will be important to assess whether other MSA variants and earlier Neandertal-associated industries of Europe and West Asia also fall within a cohesive distribution of time and space.

My inclination is to interpret these cultural distributions in terms of information exchanges. In that regard, it is essential to consider smaller units of information transfer. An entire culture is inherited by no one. A stone tool manufacturing technique, on the other hand, may be manifested in multiple artifacts and may have been learned by many individuals over thousands of years. I would be very interested in the temporal patterning within the Howieson's Poort; a question that the dates may now allow archaeologists to answer.

References:

Jacobs Z, Roberts RG, Galbraith RF, Deacon HJ, Grün R, Mackay A, Mitchell P, Vogelsang R, Wadley L. 2008. Ages for the Middle Stone Age of Southern Africa: Implications for human behavior and dispersal. Science 322:733-735. doi:10.1126/science.1162219

Hamilton MJ, Buchanan B. 2007. Spatial gradients in Clovis-age radiocarbon dates across North America suggest rapid colonization from the north. Proc Nat Acad Sci USA 104:15625-15630. doi:10.1073/pnas.0704215104

Haynes G and 14 others. 2007. Comment on "Redefining the age of Clovis: Implications for the peopling of the Americas." Science 317:320. doi:10.1126/science.1141960

Waters MR, Stafford TW, Jr. 2007. Redefining the age of Clovis: Implications for the peopling of the Americas. Science 315:1122-1126. doi:10.1126/science.1137166

This is the second in a series on information theory and tests for recent selection. The first entry, "Information theory: a short introduction" reviewed the basic concepts of information measures and their background.

The International HapMap is a massive project to determine the genotypes for up to 3 million single nucleotide polymorphisms (SNPs) in samples of people from 11 population samples around the world. The current data release (Phase 3) includes genotypes for a subset of over 1.5 million SNPs in 1,115 people. The 11 population samples include people of African ancestry from the US Southwest, Utah residents of Northern and Western European ancestry, Han Chinese from Beijing, people of Chinese ancestry from Denver, people in the Houston Gujarati Indian community, Japanese people from Tokyo, Luhya and Maasai people from Kenya, people of Mexican ancestry from Los Angeles, Italians in Tuscany, and Yoruba from Ibadan, Nigeria.

As impressive as this effort is, we may wonder why exactly SNP genotyping of so many people is a valuable enterprise in itself. The project’s homepage includes this short statement:

There are theoretical and practical objections to this simple explanation (as I discussed here last month). However, what no one involved with the project seems to have expected is the extent to which the data would demonstrate the importance of recent adaptive evolution in human populations.

Here, I am describing some of the ways that we can test hypotheses about natural selection by using the SNP genotypes from the HapMap. This is a theory-centric description, with some digression into practical aspects of handling the genotype data. First, I consider how we might use information theoretic concepts to test the hypothesis of independence between two genetic loci.

I lectured this week in my Biology of Mind course about information theory, and in particular the concept of Shannon entropy. I’ve typed up a few notes for my students, and I’m cross-posting them on my own blog because they are relevant to another topic I’ll be writing about: discovery and testing of natural selection in the human genome. You see, the kind of data that are presently being collected as part of the International HapMap , single nucleotide polymorphisms (SNPs), are naturally treated by information theoretic measures. So first, it may help to define the essential concepts of information theory.

HOMER: But I'm just one man. What can I do?

LISA: You can destroy that evil plant!

HOMER: But what can I do as an individual?

Last week, I wrote a bit about research into the affect of "gossip" in decision-making. By putting people in a situation where they played an economic game with a series of different opponents, the research found that people sometimes behaved as if short pieces of "gossip" about other players determined their decision-making much more than information about those players' past game play. In other words, a short opinion about another player from a third party was more influential than a long list of the other player's actual game history.

I wrote that it was likely a matter of priming:

The relevant point here is that the positive and negative "gossip" terms carry vastly less information than the statistics. And that's what makes them valuable. The choice before a given player is binary -- should you give your partner 1.25 or not? This choice is made vastly easier by the supply of a single bit of information -- is he a jerk, or not? You may be able to derive this information from the statistics, and if you trust someone else, you may make the wrong decision. But it's not hard to see that this experimental setup primes the players to seek a binary information source, the cost of a "wrong" judgment is pretty low, and there is no benefit in giving "wrong" advice.

Well, there's another context in which this kind of priming is important -- and it shows the same story from the opposite side. Razib points me to a paper about priming and charity-giving. In the paper, the authors -- Deborah Small and colleagues -- discover that donors respond poorly to statistical information. The abstract sums it up:

When donating to charitable causes, people do not value lives consistently. Money is often concentrated on a single victim even though more people would be helped, if resources were dispersed or spent protecting future victims. We examine the impact of deliberating about donation decisions on generosity. In a series of field experiments, we show that teaching or priming people to recognize the discrepancy in giving toward identifiable and statistical victims has perverse effects: individuals give less to identifiable victims but do not increase giving to statistical victims, resulting in an overall reduction in caring and giving. Thus, it appears that, when thinking deliberatively, people discount sympathy towards identifiable victims but fail to generate sympathy toward statistical victims.

In this experiment, the potential "donor" was either told a "Save the Children"-type story about a hungry child in Malawi, or she was told a set of statistics about the health toll of food shortages across Africa. The experimenters were interested in the extent to which personifying a crisis increased giving (the "poster child effect").

But what they discovered was surprising -- personifying a problem may increase giving, but giving statistical information depresses charitable giving, even when the problem is personified by a story about a victim:

Apparently, statistical information dampens the inclination to give to an identifiable victim. This result is consistent with the tendency to give less to an identifiable victim after learning about the discrepancy in giving. When jointly evaluating statistics and an individual victim, the cause evidently becomes less compelling. This could occur in part because statistics diminish the reliance on one’s affective reaction to the identifiable victim when making a decision (Small et al. 2007:149).

Of course, this observation is problematic if the proximate goal of charitable giving is to increase social welfare. However, if people give to make themselves feel better, then you would probably expect statistical information to reduce giving -- the main effect of statistical information being to show how small one person's contribution is compared to the size of the problem. In that vein, we should only expect huge donors like the Bill and Melinda Gates Foundation to respond positively to statistics. And indeed, Bill Gates makes malaria statistics a major focus of his public talks -- he has been swayed! For individual donors, you should expect person-to-person kinds of donation to be more effective: they match the individual effect to the same size of problem.

The commonality between this situation and the "gossip" situation is obvious: people use the available information to make decisions. When the amount of information approaches binary (hungry child bad, food good), the decision is more readily made. Introduce complicating factors, and the decision is much more complicated. Notice how those commercials never show the childrens' parents? Introducing that complicating factor provides information that would suppress giving.

In terms of charity, statistical information tends to integrate the problem more fully within a potential donor's worldview, and that's bad for giving. A dollar a day won't solve the hunger problem in Africa, but it will supply you with coffee. At least, non-Starbucks coffee. Those giant problems are the government's problem, anyway.

Stories about poster children are, naturally, propagandistic -- they provide a selected "frame" on the problem, excluding information likely to detract from giving. "Gossip" is often (but not always) propaganda on a personal level. The effect of both is to reduce information, not increase it.

I don't like the way that Small and colleagues frame their paper, by promoting charitable giving as "good" and a reduction in giving as "callousness." Failing to give charitably in response to statistical information is quite rational: people have their own problems, and when you bring their attention to the fact that their money will make a very small difference, it should be no surprise that they may prefer to keep their money! Perhaps charitable giving increases "social welfare," as the study assumes. But that assumes that the frequent resort to propaganda has no social costs.

References:

Small DA, Loewenstein G, Slovic P. 2007. Sympathy and callousness: the impact of deliberative thought on donations to identifiable and statistical victims. Organizational Behavior and Human Decision Processes 102:143-153. doi:10.1016/j.obhdp.2006.01.005

In Erika Check's Nature article on celebrity genomes, she includes a passage in which Francis Collins points out a problem with public access to private genomes:

But it's not clear that all of the genome pioneers are acting altruistically. Watson said at the Cold Spring Harbor meeting on 10 May that he has not asked either of his grown sons for permission to publish his genome sequence, which 454 has said will be publicly posted in some form. That has raised questions about the responsibility of sequenced individuals to family members who share their DNA.

"This will be a challenging question, because if you're planning to put this information in a truly open database, there are issues of risk not just to you, but to your relatives," Collins says. "Jim clearly felt those risks were not such as to cause him to take action on them."

Putting your genome information online is not only about you: it includes half the genome of each of your children, half the genome of your parents, a fourth that of your grandchildren, nieces, and nephews, and so on.

I wrote about this problem two years ago, linking to a New Scientist article that described how a young man had tracked down his biological father -- using DNA samples put online by the man's relatives.

The boy paid FamilyTreeDNA.com $289 for the service. His genetic father had never supplied his DNA to the site, but all that was needed was for someone in the same paternal line to be on file. After nine months of waiting and having agreed to have his contact details available to other clients, the boy was contacted by two men with Y chromosomes closely matching his own. The two did not know each other, but the similarity between their Y chromosomes suggested there was a 50 per cent chance that all three had the same father, grandfather or great-grandfather.

OK, so this particular situation must be pretty rare. But it is a good example of a case where a parent and child may have divergent interests with respect to genetic information. On the obvious level, the son wants to discover his father's identity while the father may want to conceal it. On the not-so-obvious level, a grandfather may want to find children that his son may have fathered, irrespective of the father's wishes. The father in question might even be dead, might have specified in his will his wishes for all sperm donations to remain private, but a grandfather can easily circumvent those wishes through the simple expedient of publicizing his DNA profile.

Families with inherited genetic conditions are already dealing with these privacy issues, such as mothers who don't want a Huntington's test and daughters who get it anyway, revealing the mother's status (my post earlier this year, referring to Amy Harmon's NY Times article). Whole-genome scans for most people will not reveal the same, tragic, level of risk, but will generate hundreds of smaller questions -- like a load of tiny skeletons-in-the-closet.

This week in Science, Collins and coauthor William Lowrance expand on the problem. Their "Policy Forum" article notes existing U.S. federal law and regulations concerning personal data and the problems that genomic information is likely to generate in the current legal context.

Until recently, most genomic research used data and biospecimens obtained fairly directly, from the data subjects themselves or clinical repositories or specialized research collections. This will continue, as it has many advantages. But now, in efforts to increase the range and quantity of data, large-scale research platforms are being built that assemble, organize, and store data, and sometimes biospecimens, and then distribute these to researchers (see figure). The advantages of such platforms, in addition to scale, are that they can be a robust staging-point for screening data quality, fostering uniformity of data format, and facilitating analysis. Some platforms accumulate data directly (as the Framingham Heart Study does); others assemble them from a variety of sources (as The Cancer Genome Atlas, the Genetic Association Information Network, and the Wellcome Trust Case Control Consortium do and U.K. Biobank will) (7). Among the design and governance issues are whether and how to de-identify the data and at what stages to conduct scientific and ethics review.

These new data flows, genomewide analyses, and novel arrangements such as the Informed Cohort scheme recently proposed by Kohane et al. (8) are relatively uncharted territory with respect to human subjects and privacy considerations. Precedent doesn't provide sufficient guidance. For example, the Human Genome and HapMap Projects have geno-typed DNA from only a few hundred carefully selected people who prospectively consented to the analysis and to open publication after thorough explanation, discussion, and community consultation. The projects have been scrutinized closely all along. But when the data relate to more people (by orders of magnitude) or to retrospective analysis of biospecimens, then for pragmatic reasons such painstaking selection, consent negotiation, and scrutiny can't generally be achieved (Lowrance and Collins 2007:600).

The article does not really arrive at any conclusions about what should be done -- Lowrance and Collins limit themselves to a fairly dry listing of potential problems and conditions leading to them. Throughout, they emphasize the reliance of the current regulations on "de-identification" -- that is, the removal of most identifying information from sequences or samples. Under today's U.S. guidelines, data that have had identifying information removed may be used quite broadly without further consideration of human subjects protections:

Construal of genomic "human subject." If data have been de-identified but include large amounts of genetic information, are the individuals still considered "human subjects"? The answer has important implications for consent, ethics review, and safeguards. McGuire and Gibbs have urged that "genomic sequencing studies should be recognized as human-subjects research and brought unambiguously under the protection of existing federal legislation" (22), but this could be unnecessarily extreme. In the United States, the Office of Human Research Protections considers that data or biospecimens collected for one purpose but then key-coded and used secondarily for research are not "individually identifiable," and therefore the research is not human-subjects research (7). This is a strong incentive to support de-identification and to de-identify data (Lowrance and Collins 2007:602).

Lowrance and Collins mention that "de-identification" is by no means as simple as applied to substantial parts of genomes, particularly when accompanied by phenotypic data such as redacted medical histories. Routine data-mining techniques would be sufficient to identify individuals within medical research studies; matching individual genome profiles to a name may be accomplished without need to match data to a "key" if the information is unique enough.

I favor the protection of individual privacy over greater research access to research data, particularly since DNA sampling and data retention by governmental agencies has become increasingly routine. In a post directly before her Personal Genome Project Q&A, Hsien-Hsien Lei wrote "Police want to collect abandoned DNA from everyone," noting that UK police will soon have authority to collect DNA with the same legal standing as trash -- if you throw it away, it's not private. We have to assume that governments will keep multiple databases of DNA barcodes for people, that these will include other personal information, and that they will be insecure. One may argue that most of the privacy threat actually comes from these other databases, and that personal genome information adds relatively little. Nevertheless, it would be better to add nothing at all, or to generate new models accentuating security.

Since I've been thinking about information theory a lot lately, I can't help but think that some kind of cryptographic solution should be applied -- so that nobody can read a person's sequence data without her private key. A person might choose to opt-in to research studies or other projects that require genotyping data, but still the sequence would be secured by encryption.

The objection to such an approach is that large-scale, long-term studies of health attributes require samples of many thousands -- even tens or hundreds of thousands of people. Today, these datasets are routinely deindividualized and dispersed around the world to researchers involved with many different projects. There is little chance of centralized control over this information after it is dispersed -- and Lowrance and Collins describe the potential problems with changing the system. With so many participants, the genotype data are a tempting target for black-hats. Any very large-scale study, in which hundreds of researchers have access to deindividualized data, there are many chances for unscrupulous researchers to steal information or put it in situations where theft by outsiders may occur.

But practices can be implemented to reduce the risk of data loss or theft. For one thing, the main reason why those studies need so many participants is because they are waiting for people to have rare adverse health events, and don't want to wait so long for results. So they really only need to know genotype data for the small group of people who have these conditions. If decryption is restricted to such small groups of study participants, the risk of unauthorized data access would be greatly reduced.

No system is perfectly safe, but in this case the agglomeration of data from thousands or millions of individuals in single databases leads to risks that scale nonlinearly with database size. So reducing the size of data chunks available to any one person may be a significant protective step.

References:

Lowrance WW, Collins FS. 2007. Identifiability in genomic research. Science 317:600-602.doi:10.1126/science.1147699

Check E. 2007. Celebrity genomes alarm researchers. Nature 447:358-359. doi:10.1038/447358a

Range, Viranyi and Huber (2007) found that dogs exhibit imititative learning:

The transmission of cultural knowledge requires learners to identify what relevant information to retain and selectively imitate when observing others' skills. Young human infants — without relying on language or theory of mind — already show evidence of this ability. If, for example, in a communicative context, a model demonstrates a head action instead of a more efficient hand action, infants imitate the head action only if the demonstrator had no good reason to do so, suggesting that their imitation is a selective, interpretative process [1]. Early sensitivity to ostensive-communicative cues and to the efficiency of goal-directed actions is thought to be a crucial prerequisite for such relevance-guided selective imitation [2]. Although this competence is thought to be human specific [2], here we show an analog capacity in the dog. In our experiment, subjects watched a demonstrator dog pulling a rod with the paw instead of the preferred mouth action. In the first group, using the “inefficient” action was justified by the model's carrying of a ball in her mouth, whereas in the second group, no constraints could explain the demonstrator's choice. In the first trial after observation, dogs imitated the nonpreferred action only in the second group. Consequently, dogs, like children, demonstrated inferential selective imitation.

Last year I posted on "imitation" in infant macaques, and noted that the term has induced a lot of confusion:

Lately, the term has been limited to cases of learning where an individual is replicating the behaviors of another individual -- not only the end result, but also all the steps that lead to that end result. But the infant "imitation" quite clearly doesn't require the kind of conceptual learning that instances of "imitation" among older juveniles and adults seems to take.

The issue remains quite confusing, although there are clarifying statements on the issue to be found in the literature. A good discussion of the concept of "imitation" as applied to social learning in particular was provided by Byrne and Russon (1998), who interpreted learning of action sequences from the perspective of hierarchization:

To explain social learning without invoking the cognitively complex concept of imitation, many learning mechanisms have been proposed. Borrowing an idea used routinely in cognitive psychology, we argue that most of these alternatives can be subsumed under a single process, priming, in which input increases the activation of stored internal representations. Imitation itself has generally been seen as a "special faculty." This has diverted much research towards the all-or-none question of whether an animal can imitate, with disappointingly inconclusive results. In the great apes, however, voluntary, learned behaviour is organized hierarchically. This means that imitation can occur at various levels, of which we single out two clearly distinct ones: the "action level," a rather detailed and linear specification of sequential acts, and the "program level," a broader description of subroutine structure and the hierarchical layout of a behavioural "program." Program level imitation is a high-level, constructive mechanism, adapted for the efficient learning of complex skills and thus not evident in the simple manipulations used to test for imitation in the laboratory. As examples, we describe the food-preparation techniques of wild mountain gorillas and the imitative behaviour of orangutans undergoing "rehabilitation" to the wild. Representing and manipulating relations between objects seems to be one basic building block in their hierarchical programs. There is evidence that great apes suffer from a stricter capacity limit than humans in the hierarchical depth of planning. We re-interpret some chimpanzee behaviour previously described as "emulation" and suggest that all great apes may be able to imitate at the program level. Action level imitation is seldom observed in great ape skill learning, and may have a largely social role, even in humans.

I think it may be valuable to add yet another hierarchical level below the "action" level; perhaps a "fine motor" level of imitation. This would be the level that fine imitation of motions like a tennis swing occupy -- people learn these through intensive practice, including slow breaking-down of the motor sequence into separate steps that are put together into a single rapid motor performance. This kind of motor imitative learning may not be present in any other animals -- in fact, I doubt there is any evidence of this in prelinguistic hominids. Perhaps the process of analysis of action at this level requires language to negotiate the time scale of modeling and the process of segmentation.

Range et al. focus on a relatively simple version of imitative learning that involves some conceptual understanding of both the goals and the steps taken while performing an action:

This type of social learning from a conspecific model clearly exceeds purely motivational and perceptual forms of social influence, such as social facilitation and stimulus enhancement [22], as already demonstrated in dogs [23]. It also deviates from simple forms of behavioral matching, such as response facilitation, i.e., the priming of an action already in the repertoire of the observer [24], because, even though the pretest showed that all animals had paw use in their repertoire, observers of both experimental groups started out by using their mouth instead of the demonstrated paw action to manipulate the rod. The quick and radical shift in the mouth-free group to adopt the paw action, for which there is no tendency in the control group, indicates an imitative form of social learning according to Thorpe [25], e.g., "as a significant elevation in the frequency of an observed action over the normal probability of its occurrence" [26] (reviewed in 27, 28, 29, 30 and 31, but see also 32 and 33).

The experiment involved an apparatus that required the dogs to pull down on a rod to obtain a food reward. Naïve dogs tended to use their mouths on the apparatus. The experimenters set up some dogs so that they could see a trained dog pull down the rod with her paw instead of her mouth. And, this is important, the dogs were primed with communicative cues from the model dog and the humans. Then, these groups were allowed to try the apparatus; upon which a high fraction of the experimental groups started using their paws to pull down the rod.

So the dogs are imitating in at least a chimpanzee-like or toddler-like way. And we bred them to do it! That may indicate that the basis for this ability is relatively shallow, in that it can be elicited with concerted selection, and without a long process of specialized evolution.

References:

Range F, Viranyi Z, Huber L. 2007. Selective imitation in domestic dogs. Curr Biol (in press) doi:10.1016/j.cub.2007.04.026

Byrne RW, Russon AE. 1998. Learning by imitation: a hierarchical approach. Behav Brain Sci 21:667-684.

Horner V, Whiten A. 2005. Causal knowledge and imitation/emulation switching in chimpanzees (Pan troglodytes) and children (Homo sapiens), Anim Cogn 8:164-181

Whiten A, Horner V, Litchfield CA, Marschall-Pescini S. 2004. How do apes ape? Learn Behav 32:36-52.

Voelkl B, Huber L. 2000. True imitation in marmosets. Anim Behav 60:195-202.

From a passage on the statistical behavior of aggregates and probability theory, p. 64-65 in Entropy for Biologists by Harold J. Morowitz, Academic Press, New York, 1970:

The notion of randomness is a very important one in physics, yet difficult to describe. (Randomness has become so significant that one of the outstanding scientific publications of recent years was a book of one million random digits.) Often a process is so complicated or we are so ignorant of the boundary conditions, or of the laws governing the process, that we are unable to predict the result of the process in any but a statistical fashion. For instance, suppose we have a collection of radioactive phosphorus atoms, P³², and take an individual atom and question how long it will take to emit an electron. Here we do not know the boundary conditions, i.e., the detailed state of the nucleus, nor do we know the exact laws coverning radioactive decay. The time can take on any value. We may obtain an aggregate of such values as is done in experiments on radioactive half-lives and deduce certain features of the collection, but we may only make probability statements about the individual atom. Randomness is in a certain sense a consequence of the ignorance of the observer, yet randomness itself displays certain properties which have been turned into powerful tools in the study of the behavior of systems of atoms.

Pp. 66-67 in Entropy for Biologists by Harold J. Morowitz, Academic Press, New York, 1970 (emphasis added):

The logic of our approach may be difficult to follow since information is not a physical quantity in the sense that mass, charge, or pressure are physical quantities. Information deals with the usefulness of a set of symbols to an observer. Since information does not measure anything physical, we are free to choose any information measure we please. The definition is therefore at first arbitrary and the choice is based on a common sense estimate of the usefulness of a set of symbols. The original definition arixing from the needs of the communications industry was, to use P. W. Bridgman's words, "of such unblushing economic tinge." What in the end turns out to be surprising is that the definition which was introduced is found to relate to the entropy concept in interesting and very fundamental ways.

The New Yorker has a nice profile of origami artist (and physicist) Robert J. Lang. My print edition of Discover had a profile of Lang earlier this year, which has become an online feature complete with illustrations.

The New Yorker story sort of leads with a dog-bites-man "well-groomed artist" story. People tend to forget that the original "Renaissance men" were skilled practical chemists (pigments) and physicists (optics) as well as painters and sculptors. There is always a long series of steps involved in the creation of art, steps that require collecting special resources, creating and manipulating technology, and bringing observations into a conceptual framework.

For this last step -- the conceptualization -- origami has lately become the province of mathematics:

He would have liked to have folded insects, but, in those years, bugs, as well as crustaceans, were still an origami impossibility. This was because no one had yet solved the problem of how to fold paper into figures with fat bodies and skinny appendages, so that most origami figures, even television characters and heads of state, still had the same basic shape as the paper cranes of nineteenth-century Japan. Then a few people around the globe had the idea that paper folding, besides being a pleasant diversion, might also have properties that could be analyzed and codified. Some started to study paper folding mathematically; others, including Lang, began devising mathematical tools to help with designing, all of which enabled the development of increasingly complex folding techniques. In 1970, no one could figure out how to make a credible-looking origami spider, but soon folders could make not just spiders but spiders of any species, with any length of leg, and cicadas with wings, and sawyer beetles with horns. For centuries, origami patterns had at most thirty steps; now they could have hundreds. And as origami became more complex it also became more practical. Scientists began applying these folding techniques to anything -- medical, electrical, optical, or nanotechnical devices, and even to strands of DNA -- that had a fixed size and shape but needed to be packed tightly and in an orderly way.

This progression is illuminating. At first, thirty steps or less -- and yet, this was a hundreds-of-years-old art form. The constraints were learning (mastering the basic techniques of folding and progressively more difficult combinations of steps) and innovation (extremely high probability of failure in objects of greater than this complexity).

Additionally, we may consider that aesthetic considerations coming from Japanese culture have driven some constraints (beauty, simplicity).

These cultural constraints themselves are process-dependent. If it were very easy to have made more complicated patterns, the value of simplicity might have been less apparent; on the other hand, the existence of an aesthetic of simplicity tends to constrain artists from experimentation with more complex forms.

Here, the artists begin with a raw material (paper) that is highly standardized is shape and form (flat, usually square) and originally, all origami was made on plain white paper. This has changed so that an endless variety of colors and textures are now used, but the starting point is still a uniform flat piece of paper.

With successive steps, there are many more ways to go wrong than to go right, and these compound as the piece becomes more complicated. Even the simplest folded crane has a high element of non-obviousness about it. No one on earth could fold an insect as recently as thirty years ago.

Simplicity may be a design feature. It may equally result from limited information -- either because information transfer is limited or because the information quantity increases too rapidly for efficient

The problem has been to devise a method for finding longer pathways to more complex shapes in an exponentially increasing sea of jumbled-looking possibilities. Lang is famous in large part because he developed software to help the process. As the Discover article makes clear, the software doesn't solve all the problems, it just narrows the possibilities. Still, this provides a light at the end of the tunnel for many complex designs -- the designer can at least know that there is a way, even if it is going to take some doing to figure it out.

At the same time, information transfer between origami artists has massively intensified. The New Yorker article lays out a short history of the art form, which blossomed during the mid-20th century as paper-folding was included in elementary school curricula, and popular magicians like Harry Houdini published books on the topic. The blossoming of more and more complex origami forms during the 1990's was made possible by international organizations and competitions -- effectively providing incentives for high-stakes exploration of folding-spaces. As people spent more time on origami, they found new ways to support themselves, from publishing instructional books to consulting with NASA on ways to fold solar panels. In short, origami became a skilled trade with multilevel professional communications.

So, if we ask the question, "How did this change in technology come about?", there are at least three answers. First, origami was adopted broadly outside the culture of its origin, causing new aesthetics to be applied to its products. Second, the incentives for innovation greatly increased, causing people to spend more time exploring folding pathways with high failure rates. Third, new search strategies and communication strategies increased the acquisition and transfer rates of information between origami artists.

With these changes, a group of artists following a broad but shallow tradition made a transition to a rich and deep field of possibilities, but a field increasingly constrained by a few important pathways widely shared among them. The essential methods for creating the narrow-appendages shared by insects, animals and human forms are a key example. Once these techniques were worked out, they were shared, subjected to subsequent alterations, and transmuted into legs, antlers, and horns.

(As an aside, the fact that magicians were an important part of this transition is characteristic of the same changes occurring in the field of illusion. The vast increases in complexity of equipment and processes, and their proliferation into more and more "unique" tricks has

How can you not love this?

In fact, origami as therapy has its proponents in 1991, at the Conference on Origami in Education and Therapy, a mental-health professional presented a paper detailing her origami work with prisoners. "The most rewarding of experiences," she wrote, "was that of observing the effect that Origami had on psychopathic killers."

(via Gene Expression)

The "Mind Matters" feature on the Scientific American blog has a commentary up by psychologist Sian Beilock. The commentary reviews last year's research that showed assigning a simple self-esteem-building essay can have a large impact on testing performance for minority students.

What I found much more interesting was Beilock's work on performance under pressure, when people are liable to "choke". I found a University of Chicago-sponsored description:

Her own experience as a lacrosse player at the University of California, San Diego, fueled Beilock's first questions about performance. To answer them, she retired her lacrosse stick and hit the putting green. Like riding a bike, Beilock says, putting becomes largely automatic once mastered, making it a "nice test bed" to gauge pressure's impact on golfers. When skilled players -- undergraduate subjects with two or more years of varsity golf experience or a PGA handicap lower than eight -- were asked to sink the ball while simultaneously identifying a specific word from a tape recording, putting ability came through unscathed, despite extra demands on concentration. Force these same experts, however, to think about their skill in a way they normally don't, such as focusing on club-swing distance or elbow position, and performance suffers. The extra attention, explains Beilock, is a common side effect of pressure situations that disrupts the flow of a well-honed activity, throwing off even the most talented individuals.

"Automatic" performance has long been fodder for intro psychology classes; it's a phenomenon that everybody can recognize. But there is a difference between the kind of automatic performance examined here -- a golf swing -- and automatic performance during other activities, like driving a car.

Driving is a long, continued activity that can occur with minimal conscious attention, at least in good conditions. But take a driver and distract him with a cell phone (or some kind of word-identifying task), and performance degrades.

So why don't the golfers experience the same performance degradation? The answer is that their "automatic" performance is in the form of highly practiced short action sequences. These short action sequences, by the way, are the kind of thing handled in the posterior part of Broca's area and the adjacent premotor cortex (link). The practiced actions do not degrade upon distraction, because the distraction does not interfere in any way with planning the sequence -- the sequence is already planned. But ask the golfer to "think about" his swing, and the sequence is brought back into question.

This kind of planned and practiced action sequence is fundamental to human imitation. If you can't parse a set of actions into some kind of sequence, you can't imitate it. This is the kind of imitation that other primates really aren't very good at. It's also probably fundamental to stone tool manufacture, since the fine control of knapping action depends on maintaining such short action sequences, and formulating longer strategies for segmenting them into a reduction sequence.

Other skills work in the opposite way. While a math whiz might perform calculations more quickly than a less-qualified classmate, successful execution still demands the expert’s dedicated focus. In contrast to a sensorimotor task like putting, many cognitive tasks call upon reserves of "working memory." It's the same type of short-term brain activity used to remember a number from the Yellow Pages long enough to make the call, and retention varies from person to person. In a low-stress situation -- Beilock’s subjects were told they were doing practice questions -- individuals who showed greater working-memory capacity did better on a challenging math task than lower-working-memory subjects. When pressure kicked in, however, these high-performers suffered the sharpest performance plunge. The discrepancy, Beilock says, suggests that individuals with high working memory may rely on complicated problem-solving techniques that naturally require more working-memory capacity than available under pressure. When anxiety begins to crowd that mental space, skilled individuals may not have enough room left over to solve the equation as quickly or successfully as usual.

Any of my students reading this after this week's exam should know that I sympathize!

This idea is very clever; I wonder if it isn't too clever. The hypothesis is that a larger working memory usually facilitiates a more complicated problem-solving method, which is more easily "crashed" by working memory "crowding". It's not obvious why stress should have this effect, though, which would seem to be highly maladaptive. Unless, "problem-solving" abilities evolved under a different context than "high-stress" situations. Maybe it is a very unusual strain of person who can think well under pressure -- or maybe an insensitivity to pressure is the real key?

Well, anyway, it seems clear that people really don't think as well under pressure, which is its own kind of distractor. And those little questions that give rise to self-doubt are some of the most powerful distractors, because they interfere with the process they reference.

A nice post on bee dancing and the bee sensory system at Neurophilosophy. From an information perspective, here we have a good example of adapting a nervous and sensory system to maximize information transfer through a given channel. In this case, the channel is defined by the frequency of bee wingbeats, particularly during the waggle dance. The critical sensory apparatus of the antennae turns out to be a simple machine for picking up these signals.

It was found that the antennae of mature worker bees are most sensitive to sounds with a frequency of between 250-300 Hz, and that the frequency and timing of the flagellar vibrations are accurately translated into the neural responses of the sensory cells in the Johnson's organ. The worker bees' hearing is therefore perfectly tuned to detect the movements of other bees, and the auditory system is ideal for listening in on the sounds made by other workers performing a dance no more than several millimetres away.

The extra twist is that worker bees differentiate their activities by age, and only older bees can hear the waggle dance at maximal efficiency. Younger bees who don't forage for food also don't have the sensitivity in the right range of frequencies.

It's a simple model with ontogenetic change in information receptivity.

References

Tsujiuchi, S., et al (2007). Dynamic range compression in the honey bee auditory system toward waggle dance sounds. PLoS One 2: e234. doi:10.1371/journal.pone.0000234.

On the subject of taboos, Susan Oyama has a discussion of taboo in The Ontogeny of Information.

The topic is canalization, and Oyama discusses the ideas of Harold Fishbein, in his book, Evolution, Development and Children's Learning. Fishbein discusses behavior as more or less genetically determined, invoking canalization as a mechanism leading to more rigid genetic control over behavior. Oyama points out that this is a backward reading of Waddington, whose concept of canalization was not about genetic control but instead more closely similar to developmental robusticity.

At length, Oyama comes to Fishbein's discussion of incest avoidance, in which he argues that the lack of mother-son incest in many primates is a biological reason why Freud was wrong about the development of the incest taboo in human societies.

Since [Fishbein] equates incest taboos, which sould seem a peculiarly human phenomenon involving certain types of forbidden conduct, with a simple absence or infrequency of behavior in other primates, one wonders why he seems to find the chimpanzee and macaque data more damning to Freud's ideas than the human pattern itself. That is, if relative frequency of various kinds of matings is given equivalent motivational significance in all the species under consideration, then infrequent mother-son copulation in humans surely counts against Freud at least as much as infrequent mother-son copulation in chimps... (Oyama 2000:112-113).

I would suggest that the point of the primate comparison is that the frequency of the behavior in humans alone is insufficient to falsify Freud's idea that the mother-son attraction is an aspect of human nature, which human societies suppress. If presumably non-taboo-bearing hominoids also lack mother-son mating, then the idea that such a desire is part of human nature would seem false on phylogenetic grounds.

But I included the first part as context for the next paragraph, which I think is fairly important:

The problem, of course, comes from impoverishing the concept of taboo, which has to do not with frequency of events per se but with their meaning. But this is precisely the kind of impoverishment that is necessary to this kind of "biological" reasoning. There is a persistent playing with levels of analysis in such treatments. Distinctions between the level of individual motivation and that of institutions or customs are ignored as the social, the cultural, and ultimately the psychological are collapsed to the biological. At the same time, a new level is created -- a phantom plane of genetic reality, which is not observed as such but deduced (or assumed). From this perspective, the observation of living, breathing animal-machines is of value only insofar as it gives access to the ghostly forms and causal agencies within them (Oyama 2000:113).

It seems to me that the real problem is identifying appropriate units of reductionism for cultural entities and behaviors. For the most part, biologists, psychologists, and anthropologists are all alike in treating "culture" as a single level phenomenon interpolated between "human biology" (or "human nature") and human behavior.

This leads to two interpretive approaches. In one, anything that is manifested in behavior that is not easily explicable in terms of genetic adaptation is consigned to "culture". In this perspective, culture is the inexplicable residue of biological evolution. Moreover, it is viewed as coming complete with its own evolutionary system -- a full epicycle upon the biological evolution that governs everything else. In this interpretation, culture is a bit like "consciousness" -- humans may be adapted to it, but we have a lot of trouble saying just what "it" is or how that adaptation has come about. But its essence is informational: culture is a layer of information that comes between genes and behavior.

The other approach holds that culture just is behavior. To be sure, not every behavior qualifies; the behavior has to be patterned among individuals in certain ways to be "cultural". But it exists just to the extent that individuals can perceive and act on patterns in behavior. The individual's role in culture is purely reactive in this view: individuals behave in cultural ways because their environment is patterned in cultural ways, not because they necessarily have special genetic adaptations to culture. We might term this a "constructivist" view, in that the individual's behavior is constructed by cultural patterns, rather than being merely conditioned on cultural information. Culture is a single layer in this view also -- a layer of environment that individuals perceive and act within.

Both these accounts of culture are essentially similar. They differ as to the locus of culture -- does it exist within individuals or societies? But they agree about the structure of culture -- it exists as a "complex whole". Thus, they do not lend themselves easily to reductionism, explaining in part why holistic interpretation is applied so broadly within cultural analysis.

It is within this context that the "meaning" of taboos has such importance. The "meaning" implies certain obligations on the part of individuals who detect violation of taboos. It implicates those individuals in enforcement not only because the taboo behavior is amoral, but because the taboo explicitly links morality to other kinds of undesirable or repugnant natural and supernatural consequences. Explaining a taboo against incest avoidance is at a different level than explaining incest avoidance itself.

The problem is that taboos and other cultural phenomena are themselves generally confined to a single "phantom plane" of explanation. Cultural phenomena and entities must exist insofar as they clearly shape behavior. We generally recognize that they have complex histories that may themselves be intrinsically interesting. But their role in shaping the development of individual behaviors is formless -- individuals either pass through a "cultural filter" or they take in "cultural information", but in either case all of culture is drunk from the same well.

References:

Oyama S. 2000. The Ontogeny of Information. Second edition. Duke University Press, Durham NC. Amazon

From the New Scientist technology blog (via Slashdot):

You may remember the story of some cheeky MIT students who wrote a computer programme to generate scientific papers. Well, now some researchers at the Indiana University School of Informatics have come up with an Inauthentic Paper Detector to foil it.

Mehmet Dalkilic, a data mining expert explains how it works: "We believe that there are subtle, short- and long-range word or even word string repetitions that exist in human texts, but not in many classes of computer-generated texts that can be used to discriminate based on meaning."

What is interesting is that these "subtle long-range repetitions" are definitely part of our comprehension of a text, but we don't necessarily have the confidence to claim a text is fake if it lacks them. We have the statistical sense innately that the computer in this program is making explicit.

It is one of the many ways that we help ourselves to make language more comprehensible -- a certain redundancy that keys the mind back to the subject at hand. A good writer uses those repeated phrases to make the text more understandable.

And it is one of the many reasons why natural texts have a relatively low information content, at least for their length -- they consistently follow certain patterns. For our minds, that's a good thing! It lets us understand them.

On the basis of a couple of student questions, I think it's worthwhile to reflect a bit on where I am going with this (also possibly made more clear in this post, which covers learning more extensively.

By definition, material culture is culture. From this, it follows that explaining regularities within (and among) stone tool industries is explaining cultural regularities.

On the other hand, there is much more to culture than material culture. Material culture presents one facies of culture, but other aspects may have little to do with the material remains of tool use and manufacture. By extending our consideration of archaeology to the entire chaine opératoire, from the acquisition of raw materials, the process of learning manufacturing skills, and the ultimate use, reuse, and discard of the tools, we may hope that one or more aspects of tool-related behavior will at least touch on many other elements of social life. But still, tools may misrepresent the substance of culture as a whole.

The problems understanding the whole culture at the Acheulean-MP transition are expressed by Belfer-Cohen and Goren-Inbar (1994:145-146, emphasis mine):

Of the 2 my and more, which are assigned to the Lower Paleolithic, over 1.5 my are attributed to the Acheulean Industrial Complex (Isaac 1977,1986). The immensity of this time-span -- which is quite beyond our comprehension -- can be 'blamed' for our difficultures in assessing and interpreting various phenomena observed in the material remains from this period. Prehistoric research during the last few decades has focused primarily on issues such as hunting versus scavenging, mobility, camping, division of labour and sharing, etc. Thus the study of lithic assemblages has become of secondary and minor importance. For a while it seemed as if the lithic component constituting the majority of the retrieved material remains could not contribute any information towards resolving the issues detailed above. However, new approaches to the study of lithics, employing various methods of technical analysis and procedures such as multivariate analyses (Karlin et al. 1991) have led to a better understanding of lithic manufacturing processes (Boëda et al. 1990; Pelegrin 1993). The large body of data acumulated over the long term demonstrates that, through well-planned research, lithic assemblages can provide significant insights into the cultural complexity of prehistoric people (Gowlett 1984; Pelegrin et al. 1988; Pelegrin 1990; Wynn 1993a). Nevertheless, some researchers (usually those far removed from lithic studies) still demonstrate simplistic attitudes, such as referring to a single artefact type, or a single technological feature, as if it faithfully reflected the total material complexity of a Lower Palaeolithic culture. Thus Foley (1987) treated the biface as if it embodied the entire Acheulean tradition, and as such he opposed it to manifestations of Levallois technique, which he regarded as the marker of the Middle Palaeolithic 'Levallois-Mousterian' (p. 386). But not only do most Levantine Acheulean assemblages contain a Levallois component (Gilead 1970; Goren-Inbar in press, and references therein) -- which from the outset pulls the ground from under Foley's argument -- but the paper contrasts two entirely dissimilar, basically incomparable entities -- a specific tool-type and a particular production technique. Most important, the selection of a single component as the all-encompassing representative of a complex techno-typological tradition hinders the serious search for meaningful patterning.

The "new approaches" mentioned here would be part of the "technological approach" discussed by Chazan (1997), reviewed in another post.

After a due consideration of the technical aspects of the Lower Paleolithic industries at 'Ubeidiya and Gesher Benot Ya'aqov, Belfer-Cohen and Goren-Inbar (1994) concluded that the full technical repertoire of Acheulean people was complex and advanced. Quoting Pelegrin (1993), they write:

It is of interest to note that Pelegrin (1993), in his detailed discussion of lithic production, refers to the process of biface manufacture as complex stone knapping. It involves '"a coherent critical approach" to the situation. It is also characterized by the construction of techno-morphological mental hypotheses and their evaoluation through the double filter of what is desirable and what is possible' (pp. 310-311). Summing up his observations, Pelegrin states that 'This operational competence is comparable with that of modern man in this technical context' (p. 313) (Belfer-Cohen and Goren-Inbar 1994:152).

They draw this to an interesting conclusion:

Perhaps we should not be surprised by the complexity observed in early lithic production. Applying the rules of 'General System Theory' (Bertalanffy 1968), we can compare the pattern discerned in the evolution of the lithic material cultures with that learned from Cambrian fossils about the history of life's evolution on earth (Gould 1989): 'The maximum range of anatomical possibilities arises with the first rush of diversification. Later history is a tale of restriction, as most of these early experiments succumb and life settles down to generating endless variants upon a few surviving models' (p. 47). In the history of the evolution of lithic production, too, we have the initial appearance of several production modes, used in a distinct fashion for the manufacturing of discrete morphotypes. Then, in later stages, for example the Mousterian cultures of the Middle Paleolithic, we observe a reduction in the number of production sequences and an increase in typological diversity (ibid:153).

I view the situation rather differently. The maintenance of a cultural tradition of tool manufacture requires individuals to receive both the information of how to make the tools and the motivation to learn and make the same types of tools by the same process. Some information is acquired socially -- by watching others make and use the tools. Some information is acquired physically -- by attempting to flake stone and learning its fracture properties, for example. Likewise, some motivation is physical -- the need to cut something, while some motivation is social -- the need to produce tools that resemble the group's tradition, or that carry forward the signature of a teacher, for example.

Without social motivations to maintain a tradition, individuals may have no particular reason to be receptive to the information that would allow the replication -- information concerning stereotypical reduction sequences and their outcomes. Physical motivations alone -- the need to cut something -- require relatively little in the way of social information.

In a statistical analogy, the Lower Paleolithic may usually be interpreted beyond the number of degrees of freedom it actually presents. Bifaces, core tools, and flakes made on different raw materials and with different reduction modes are not necessarily evidence of complexity. Certainly a modern-day knapper asked to recreate each of the many combinations of tools and mode in an assemblage would have a complex task ahead of him. But thirty novice knappers asked to make each of several tools and given time to work with the materials would very likely come up with many different methods. In essence, our analyses might like to make some conclusion about "complexity", perhaps for interpreting cognitive prowess, but "complexity" is simply not a parameter relevant to the assemblages

The Middle and Upper Paleolithic expresses some strong differences in tool types and manufacturing techniques across assemblages, along with greater within-assemblage consistency in technique. These two observations -- variability between and consistency within -- are sides of the same coin, since consistent differences between samples are impossible without consistent similarities within samples. They both arose because individuals were making more use of social information sources as they learned to make and use tools. Traditions are necessarily less recognizable when individuals acquire and use less social information in tool manufacture. Thus, more social information must lead to more recognizable and distinctive industries.

Continuing the statistical analogy, the later industries present more degrees of freedom. There is more information to be explained in their production techniques, and the techniques themselves explain more about the lifeways and cognition of the makers. But what is to be learned is not mainly spatial or technical abilities, but instead attention to social information and motivations for maintaining social traditions.

The point to the analogy is that the earlier tool traditions can say little about cultural systems, because the people show little evidence of having attended to cultural information in the context of tool manufacture. Even among chimpanzees, tool manufacture requires some cultural information -- observing other individuals using tools, for example -- and early humans must have had more social information about tool manufacture than this.

And some aspects of early human tool manufacture do reflect social information. For example, raw material utilization clearly varies among sites:

Figures 2 and 3 list the major tool types with the raw materials of which they were most frequently made (as can be seen in this inventory, the bifaces represent but one of several production sequences encountered at the site). Similar associations of specific kinds of raw material with specific tool types have been repeatedly identified elsewhere, especially in East Africa (Feblot-Augustins 1990 and references therein). In our opinion, these associations, which were found to be spatially and temporally transgressive at Ubeidiya, clearly attest to a continuous and effective process of information transmission. Such preference for certain selected raw materials within a given assemblage is a distinctive feature of the Acheulean, and can be viewed as indicating awareness of one or more of the following environmental potential, lithological properties or procurement logistics. Interestingly, the relative frequency of certain tool types does not seem to have been affected by the degree of availability of specific raw materials. Thus, while the scarcity at the site of limestone, for example, does not seem to have deterred the production of limestone spheroids and sub-spheroids, the abundance of basalt pebbles did not result in a higher production rate of basalt bifaces (ibid.: 150).

But the use of social information did not dominate early tool manufacture. This means that the same functional items may have been repeatedly reinvented independently. A diversity of forms and methods reflects the primary use of nonsocial information in the learning and production of tools. Each individual worked with stone, probably upon observing others do so and use the tools that resulted. But they did not attend to the specific procedures used by other individuals, at least not beyond a point. Instead, each individual discovered much about flaking techniques and the physical properties of stone by himself (or herself). That repeated trial-and-error learning spread across many individuals explains the variation in reduction sequence in early tool traditions, the relative simplicity of end results, the use of a similar range of procedures on multiple raw materials, and the widespread lack of interregional or local variations. Some variations did exist and were persistent (i.e., East Asian lack of handaxes, woodland vs. nonwoodland in Europe). These may have reflected functional considerations oor persistent and widespread differences in raw materials (i.e. bamboo). But the East Asian case in particular begs for a solution that involves social as well as functional considerations, since bifaces to occasionally occur and might presumably have been useful for a range of functions that bamboo and simple choppers and flakes could not manage.

In this case, there is some reason to see the biface not only as a functional implement but as an icon of its own manufacture. Social information is present in the form of the biface itself, to the extent that it is clearly recognizable as a product of human manufacture. A person might discover the way to replicate a biface, and even the motivation to do so if recurrently exposed to the shape and manufacture of them. Thus, the Achuelean tradition carries persistent social information, even if the ability of individuals to incorporate this social information was limited to the end product and basic fact of manufacture.

It is tempting to interpret uniformity as requiring the repeated emphasis of certain kinds of information. But this interpretation assumes a human context in which many other kinds of social information are easily received and compete for attention. Students put their names repeatedly on assignments because of long reinforcement and adverse consequences for failure. This piece of information is emphasized to drown out the noise of other social information.

But standardization and uniformity may equally be the result of limits to information transfer. We may imagine a society in which the only information that can be written is a name. In this context, it would be no mystery why there should be many objects that have names on them -- indeed, it is the only possible outcome. The biface may have emerged from a society much more like this one -- it was repeated because of the lack of other information, not because of a persistent reinforcement.

As an aside, individuals who use lots of social information in their toolmaking procedures -- essentially, who learn to make tools by observing the precise steps that others use to make tools -- may not always imply that groups will arrive at different toolmaking traditions. There may be, after all, one best way to make tools, and different groups may arrive at similar methods semi-independently. But the observation that later technologies appear to become more consistent in their use of reduction sequences is strong evidence for the greater employment of social information irrespective of the appearance of regional or local variations. These variations may be a welcome confirmation of the model, but the degree of such variations cannot be predicted from it.

References:

Belfer-Cohen A, Goren-Inbar N. 1994. Cognition and communication in the Levantine Lower Paleolithic. World Archaeol 26:144-157.

Noble W, Davidson I. 1996. Human evolution, language and mind. Cambridge University Press, Cambridge UK. Amazon

That's "Upper Paleolithic", not "Upper Peninsula." Barton et al. (1994) discuss the interpretation of Paleolithic art in Western Europe. A good short summary is this passage (p. 190):

Viewing art as a communication medium that monitors information flow allows us to propose an explanation for observed patterns in Palaeolithic and post-Palaeolithic art, and to model changing alliance networks in European forager populations from the late Pleistocene through the early Holocene.

They conclude that the tradition of parietal art in Southwest France and northern Spain reflects demographic pressures associated with increases in population density, stress on existing alliance networks, and claims for property rights.

The information exchange approach taken here argues that parietal art is an example of Wiessner's emblemic style; that emblemic style is, among other things, a monitor of demographic stress, and that the appearance of parietal art in late Pleistocene Europe resulted from the closing of social networks under increasing population density (ibid:199).

They suggest that this use of emblemic style is merely an archaeologically visible instance of a system based in archaeologically invisible referents, and that this visibility is a function of group size and resource availability:

As noted, both assertive and emblemic style almost certainly existed among prehistoric foragers, but are not usually recoverable archaeologically. Ethnographic data indicate that emblemic style in forager contexts is often associated with features of the landscale (e.g. Denbow 1984; Lewis-Williams 1984). That is, it functions to identify sacred localities, prominent topographic features, the boundaries of more or less exclusive economic territories and other geographic landmarks. Under conditions of low population density, changing group membership and open social networks, the significance of such landscape features is transmitted from one generation to the next by means of the oral traditions of small, fluid social units; the physical marking of such features is rarely necessary. With aggregation, however, the need for more effective means of both inter- and intra-group communication arises. It becomes necessary to reinforce oral tradition within larger social units, whose members might not participate in a single tradition nor interact with one another on a regular basis.

I wanted to note the contrast they draw between two "explanatory paradigms" for patterns in Paleolithic art:

Art and social interaction

The first sees distributional patterning in Paleolithic art, and more generally style, as a monitor of the degree of cultural affinity among social groups through time or in terms of shared cultural traditions. This 'social interaction' model has roots in both the Old and New Worlds. In North America, it dates back to A. L. Kroeber, Clark Wissler and the 'culture area' studies of the 1930s. Social interaction theory defined style in a strongly normative way as repetitive behavior that acts as a kind of psychological 'filter' to constrain variety and reduce information overload. Style functions at the level of the individual and is essentially passive; that is, it reflects normative constraints learned unconsciously through enculturation. In art and in artefact design, style is construed to exhibit modal properties taken to reflect, an a more or less direct way, group norms and values. These modal properties are often considered to be isomorphic with the temporal and spatial boundaries of identity-conscious social units -- in other words, an ideational definition of style, but with alleged material correlates (Clark 1993) (ibid.:186).

In other words, "style" in artifacts reflects the boundaries of meaning in the minds of people who made the artifacts. It is unintentional -- a byproduct of the process of learning a limited set of information. Because of a lack of intentionality, the analysis of style may uncover the social units themselves, consisting of individuals who shared boundaries of meaning. Style is therefore a marker of cultural affinity.

Art and information exchange

The second approach to art is essentially a functionalist one that views it as the remains of communication systems involving the exchange of information (Braun and Plog 1982). As employed here, the 'information exchange' theory of style originated with Polly Wiessner (1983, 1984, 1985), who views art as an act of social communication defined, at various levels and scales, by style. Style, in turn, as its behavioral basis in a fundamental human cognitive process: the personal and social identification of images through visual comparison. In sharp contrast to the pattern searches of social interaction theory, style is defined here by its determining processes, rather than by its material conditions.

In this paradigm, "style" is a functional element of artifacts that may or may not correspond to cultural entities. It has an adaptive function in the intentional transfer of information, such as information about social status or property. Since the imposition of style is deliberate, it does not mark cultural units in a straightforward way, but instead marks patterns of activity that may or may not be linked to cultural entities.

I like this second concept, because it assumes that people are intentional agents with respect to stylistic representation. But I think that elements of the first are also necessary, since there clearly are boundaries to meaning that are important in the function and interpretation of stylistic information. Indeed, without boundaries to delimit meaningful representations, there can be no transfer of information (The importance of these delimitations on meaning is well illustrated by modern art, which is often more about the placement of boundaries to meanings than it is about stylistic or representational elements).

References:

Barton CM, Clark GA, Cohen AE. 1994. Art as information: explaining Upper Palaeolithic art in Western Europe. World Archaeol. 26:185-207.

An interesting story from Howard Hughes Medical Institute (via Science Blog) about the information content of whale song. They don't know what the whales are communicating, but they can assess the number of bits transmitted:

[HHMI predoctoral fellow Ryuji] Suzuki said that information theory also enabled the researchers to determine how much information can be conveyed in a whale song. Despite the "human-like" use of hierarchical syntax to communicate, Suzuki and his colleagues found that whale songs convey less than one bit of information per second. By comparison, humans speaking English generate 10 bits of information for each word spoken. "Although whale song is nothing like human language, I wouldn't be surprised if some marine mammals have the ability to communicate in a complex way," said Suzuki. "Given that the underwater environment is very different from our world, it is not surprising that they would communicate in rather a different way from land mammals."

There seems to be much emphasis on the "hierarchical" aspect of the songs, and this is important -- a single call is made up of many smaller subunits, each of which may carry information content and the arrangement of them may itself carry information content.

Suzuki, who began the project as an electrical engineering undergraduate at the University of Massachusetts, Dartmouth, worked with Buck and Tyack to develop a computer program to break down the elements of the whale's song and assign an abstract symbol to each of those elements. Suzuki wanted to see if he could design a computer program that enabled scientists to classify the structure of the whales' songs.

He used the program to analyze structural characteristics of the humpback songs recorded in Hawaii. To measure a song's complexity, Suzuki analyzed the average amount of information conveyed per symbol. He then asked human observers who had no previous knowledge of the structure of the whale songs to classify them in terms of complexity, redundancy, and predictability. The computer-generated model and the human observers agreed that the songs are hierarchical, confirming a theory first proposed by biologists Roger Payne and Scott McVay in 1971.

...

The structure of the humpback whale song is repetitive and rigid. The whales repeat unique phrases made up of short and long segments to craft a song. There are multiple layers, or scales, of repetition, denoted as periodicities. One scale is made up of six units, while a longer one consists of 180-400 units. The combined periodicities give the song its hierarchical structure.

A hierarchical format is vastly more learnable than any nonhierarchical alternative capable of encoding an equivalent amount of information, so it should not be surprising that this structure would have arisen in another highly communicative species. It emphasizes that limits on information transfer are just as fundamental to the evolution of social intelligence as limits on optics are to visual perception.

CNET is running a series of articles on the kind of intelligence required for the world of changing technology. The first installment starts thusly:

Today, terabytes of easily accessed data, always-on Internet connectivity, and lightning-fast search engines are profoundly changing the way people gather information. But the age-old question remains: Is technology making us smarter? Or are we lazily reliant on computers, and, well, dumber than we used to be?

The article's answer is that different skills don't mean different reasoning and learning. Not unexpected, since business' focus in the wake of technological change is always training and retraining the same minds for different skillsets.

The main idea is how memory is less necessary when you have devices to keep track of things for you. I suppose if Sherlock Holmes' theory of mind is right, that means we should be able to fill up our minds with deeper thoughts:

"It's true we don't remember anything anymore, but we don't need to," said [Jeff] Hawkins, the co-founder of Palm Computing and author of a book called "On Intelligence."

"We might one day sit around and reminisce about having to remember phone numbers, but it's not a bad thing. It frees us up to think about other things. The brain has a limited capacity, if you give it high-level tools, it will work on high-level problems," he said.

Of course, this presupposes that the brain isn't full of cognitive adaptations that now lie fallow and useless in today's high-tech world. Or get filled with videogames and movies. I guess these fall under the Everything Bad Is Good for You theory.

I wonder what you would call a specialized cognitive adaptation that could be readily reprogrammed in different cultures for different purposes?

Via Instapundit, a link to a review of the new book Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. Looks like a very interesting book, covering topics ranging from why teachers and sumo wrestlers both cheat, to the economics of drug dealing.

On this latter subject, the review paraphrases the book's argument:

A chapter titled "Why Do Drug Dealers Still Live with Their Moms?" provides an analysis of the economic workings of a Chicago crack gang, based on data collected (initially at high personal risk) by a young sociologist named Sudhir Venkatesh. The upshot is that the crack trade, even at its market peak, was lucrative only for those at the top of a selling organization. The gang's foot soldiers made less than minimum wage and faced a 1-in-4 risk of being killed over four years. (In the same time, being a timber cutter, the most dangerous legitimate job in the U.S., carried a 1-in-200 risk.) These drug dealers struggled desperately to reach the gang's upper echelons, but few would make it.

From Instapundit:

My historian-brother often says that one of the most interesting phenomena that he's observed is the cross-cultural willingness of people to trade away economic benefits for status. I suspect that this is one example of that. So, in a surprisingly similar way, is being a politician. That's an obviously poor economic move for most folks. But one of the drug dealers in Price's book talks about how he likes the way he becomes the center of attention when he enters a room full of junkies. Politicians, I think, get the same thing, especially in the bubble-environments of Washington, or state capitals. I suspect, in fact, that people are, to varying degrees, hardwired to get an endorphin rush from that sort of attention, just as they're hardwired in varying degrees to respond to drugs.

As Reynold's readers comment, you can replace "crack" and "dealers" with "universities" and "professors" and come to a similar conclusion. People rarely maximize their monetary gains with their choice of careers or activities. After all, any of us could take an extra job at Taco Bell to turn our downtime into cash. If we were Scrooge McDuck-like packrats who did nothing but fill bank accounts, any extra dollar (or the employee burrito discount) would impel us. The fact is, people don't lust for money in this way. "The pursuit of happiness" is what most of us are spending most of our time planning, if not actually pursuing.

The thing that gets me is why anyone should expect that humans would be rational maximizers with respect to money, when money is a very recent cultural innovation. Status and prestige are much older influences on human behavior. Indeed, rank is a central component of many mammalian (even vertebrate!) social groups, and "status" is just another way of saying "rank" in human cultural terms. Determining rank must be an important part of the mental calculus of any social species, and humans if anything are beyond the capabilities of most. So there is every reason to think that humans should be adapted not to economic maximization, but instead to status enhancement -- mainly by seeking prestige from other people.

Sure, it is true that a lot of money can buy prestige and status of a certain kind. And if you just looove the feel of 1000-thread-count sheets and the looks you get when you drive an Astin-Martin, this kind of status may be exactly what you are looking for. But the pathways to status today are as varied as our lifestyles and interests. Consider the rewards that come with being the best-dressed Dr. Frank-N-Furter at the local Rocky Horror Picture Show revival. Or that comes along with the publication of an academic book. These things lead to prestige and status -- at least within a fairly loosely defined but circumscribed group of people -- but they certainly don't lead to economic benefit (when compared to the opportunity cost). Present-day human societies provide many channels for status-seeking, many of them non-overlapping. The effect is that most people can channel their activity into patterns that result in prestige-enhancement from some group of peers, at least sometime during their lives. And people for whom such courses are not available often face psychological consequences such as depression.

I think we can safely speculate that prehistoric human societies were the same as living ones in this respect. On one hand, population sizes and densities were smaller in the distant past, so that the same spectrum of options for prestige-seeking were certainly not available. That is to say, if your dream is to be a writer for the Dick Van Dyke Show and you live in Java around 200,000 BC, you're pretty much out of luck (unless you know Rose Marie's agent). But on the other hand, smaller population sizes mean fewer competitors for the range of behavioral specializations available. Less competition means more prestige for the same degree of talent, productivity, or knowledge.

Like societies in other primate species, such as chimpanzees, humans do not compete for status as a zero-sum game. Status is a product of a complex network of social interactions, and its cumulative effect is not easily predictable from the interactions themselves. This makes it a difficult calculation, and therefore leads to the hypothesis that our minds -- along with the minds of other social species -- possess special adaptations to perform such calculations. The adaptive value of such mental functions would be to shape decisions in a way that tends to increase status, and thereby fitness to the extent the two are correlated. Loosely put, this is the "social brain hypothesis," described by Robin Dunbar (1998; 2003).

There is, however, an important distinction to be made. The "social brain hypothesis" was proposed to explain brain size -- predicting that more social species will have larger brains for their body size.

[P]arsimony and biological common sense would suggest that it is group size that drives brain size evolution rather than brain size driving group size and that group size itself is a response to an ecological problem (most probably predation risk (van Schaik 1983, Dunbar 1988, Hill & Dunbar 1998)). Although the hypothesis has been tested by determining how neocortex volume constrains group size and other social indices, the evolutionary logic is that the need to maintain coherent groups of a particular size has driven neocortex volume evolution through its demands on cognitive competences. The most succinct and parsimonious causal sequence with fewest unsupported assumptions is that the window of opportunity provided for more intensely bonded social groups and the social skills that underpin this was the crucial selection pressure for the evolution of large brains, even though simple ecological pressures (e.g., the shift to a more frugivorous diet) may have been instrumental in kicking off the process. In these terms, any associated ecological skills may be seen as the outcome of the opportunity provided by an increase in general purpose intelligence generated off the back of the social requirements. To argue the reverse sequence (that large social groups are a by-product of having evolved large brains to solve simple ecological problems) is, as with the various ontogenetic hypotheses, to leave unanswered the problem of the costs of social living (Dunbar 2003:169).

If the relationship between social group size and brain size--or more specifically, neocortex size--could be quantified, then it would be possible to predict the group size of an extinct species based on the size of its brain. This is precisely what Dunbar has attempted for fossil hominids, and for living humans. Such an estimate for a population or species has become known as the "Dunbar number", and has found application in areas of the social sciences beyond animal behavior (most interestingly, in predicting characteristics of clandestine terrorist networks based on the kind of communication possible between the members).

This trend appears to hold across large taxonomic groups, like the primates. But within family level taxa, like the hominoids, it is unclear how much predictive power the hypothesis may have. In particular, the small social groups of orangutans appear to be a poor fit to the model when compared to their brains, which are similar in size to those of chimpanzees. So in the effort to predict characteristics of past species of hominids based on the size of their brains, the logic of the "social brain hypothesis" faces some perhaps insuperable problems.

I'm more interested in the aspects of mental flexibility and variability that arise as a consequence of selection for increased social tracking. As social groups grow in size, the number of binary interactions that any one individual needs to trace increases exponentially. The focus on this number of interactions is what drives the Dunbar model, along with the expectation that each of those interactions may require the investment of energy to maintain social status. But an increasing number of interactions is not the only change associated with larger social groups. We may also expect an increase in the number of kinds of interactions. This is a qualitative change that accompanies the quantitative change resulting from increasing population size.

Human evolution may or may not have seen the an increase in group size compared to our Miocene ancestors. Chimpanzee communities today number somewhat larger than human hunter-gatherer groups, on average. But chimpanzee committees themselves are comprised of smaller groups of individuals who may rarely see each other and the course of a month or longer. And human hunter-gatherers may coalesce into larger groups for social purposes. The complexity of social interactions within each of these species is beyond that of most other social species, factoring the variation in time spent with other individuals and the gradation of possible interactions from cooperation to aggression to reconciliation. Human evolution may not have seen an increase in group size, but it certainly saw an increase in the flexibility of interactions with individuals who were members of other groups. It also saw an increase in the degree of behavioral differentiation of individuals within groups. There were without doubt knowledge specialists for each of the components of ancient human cultures. This probably included specialists in tool manufacture, in hunting, in plant foraging, in midwifery, in negotiation with other groups, in knowledge of the landscape, and many other realms.

The possibility of specialists in different types of knowledge creates an "Information economy" in which people may attain and secure status on the basis of what they know about a specialized activity. If people can monopolize special information, then they can make themselves indispensable. And where different kinds of information are indispensable, different group members can leverage a more egalitarian social role.

To see this system in action, one need look no farther than the plot of any good story. Take for instance Homer's Iliad. In the descriptions of the war planning and decision-making, we see a very small subset of an ancient society--only the warrior class of ancient Greece. But even within this limited set, different individuals have different attributes that are essential to the enterprise. Agamemnon brings political power and finesse, Achilles brings martial skill and the intense loyalty of his followers, Odysseus brings clever planning and deceit, and Nestor brings a measured eye for tradition and restraint. Is it possible that in this microcosm of the story we can see the interactions within human social groups since the dawn of humanity? Clearly at the least it illustrates the ways that humans angle toward the unique role in their societies, that may defy ready quantification.

Returning to the issue of money, clearly economic advantage is one of the ways that we compete for social status today. Perhaps the majority of interactions that we have with other individuals have to do with the exchange of currency, or the social expectations that come from labor, consumerism, or other economic relations. And money even becomes part of our close social relationships, between family and friends. But few of us define ourselves in terms of money. Most people find other things that are more important to them. The discussion of money itself is something of a social taboo in our society, with most people considering it ill-mannered to directly raise the issue of how much someone makes, or how much they paid for something valuable. We play games of status around the issue of money, avoiding it almost whenever possible.

These games are an intricate result of our mental adaptations. The quest for social prestige and status is played on the field where the rules are fluid. Today's rules are different than those of the Stone Age, but the way we count our chits is mostly the same.

References:

Dunbar RIM. 1998. The social brain hypothesis. Evol Anthropol 6:178-190.

Dunbar RIM. 2003. The social brain: mind, language, and society in evolutionary perspective. Annu Rev Anthropol 32:163-181.