Saturday, May 27, 2017

Cromwell, Superman

A brief review of Wolf Hall.

The press of business prevents a longer post, so just a few words on a novel I have been laboring over, Wolf Hall, by Hillary Mantel. First, the style- it is offensively eliptical and unclear, not indicating dialog or who is speaking or thinking. How it could have been the rage of the prize committees and reviewers a few years ago is simply beyond me. Edgy, yes. But respectful of the reader, no. Comparing it to another bit of current reading, Thomas Hardy's Far From the Maddening Crowd, couldn't make the contrast starker. Hardy treats his reader as well as his subjects with great respect. Humor, yes, and deep insight, but first of all clarity, and rich context and exposition. 

Ann Boleyn, secondary subject of Wolf Hall.

Mantel, in contrast, plays games with the reader, telling the story through a straw of cynical mannerism, hiding far more than she tells. Thomas Cromwell, the main character, is made out to be a 100% pargon of competence and compassion, and all other characters are given temperatures directly in proportion to how close they are to him. Those he hates are vile, those he likes are good and virtuous. Yet the story is not really told from his perspective, but from Mantel's snarky omnicient voice, making all this characterization absurd as well as historically unbelievable. I assume that the BBC production was able to, thankfully, jettison virtually all the novelistic apparatus and return to whatever of the history was presentable, using only the broadest outlines of Matel's selection of scenes and personalities.

Sunday, May 21, 2017

Sponges Are Animals Too

A brief look at some innovations at the bottom of the animal evolutionary tree.

Of the many great innovations in evolution, the development of multicellular animals ranks pretty high. It was based on the many innovations that had previously generated eukaryotes and their ramifications into protists and other complex single cells. It was a big step to realize (unconsciously!) that, however big and complex one could make a lone cell, (think of a paramecium), possibilities of greater scale and specialized organization were being left on the table. Sponges are the most primitive animal that currently exists. To us, sponges are small and simple. But to their protist brethren, they are giants- vacuum cleaners of the sea that suck up protists and other detritus like so many dust mites.

Sponges have several distinct cell types, like outer epidermal cells, pore cells, and interior collar cells that have the flagella that keep water moving through the pores. Sponges also contain wandering cells that secrete collagen in the intersitial areas- the same protein that keeps our bodies together as well. And they have primitive muscle cells / myocytes, which can cause contractions to close pores when needed, and also the larger openings of some species. These myocytes necessarily have some electrical communication, to coordinate their activities around the osculum. So, not exactly Brad Pitt, but they have quite a bit of anatomical complexity

A sponge.

Key innovations of this (relatively) massive and complex organism were not just the novel molecules, such as collagen, and skeletal "spicule" molecules, but the regulatory apparatus that generates different cell types and keeps them at their work, diligently pumping, or waving, or opening/closing, etc. A recent paper looked at this regulatory system and found that sponges already have most of the special mechanisms that other animals use to specify cell fates and body organization.

Those mechanisms revolve around gene control. In the human genome, about 1% of the DNA codes for proteins- the actual stuff of our bodies. Roughly about 7 to 10% of the genome functions in other ways, principally as regulatory regions, RNA-encoding genes, etc. The rest seems to be junk, more or less. The sponge that has been sequenced to date has a genome with roughly the same number of genes as ours (20,000), in 1/20 the genome size. So there is quite a bit less junk and complexity, but still a very large tool chest.

Particularly, it has plenty of regulatory control, and the researchers explore its use of one hallmark of metazoan gene control- histone modification. While the basic role of histones is very simple- to act like tiny wheels around which the DNA wraps, keeping it both secure and compact- its position also allows it to regulate the access to that DNA by other proteins, especially regulators of expression. Higher metazoans have a dizzyingly complex "histone code", composed of chemical modifications to the positively charged tails. These modifications include methylation, di- and tri-methylation, and acetylation, on lysines at positions 4, 5, 9, 14, 20, 27, 79, and 122 of one of three different histones. Each modification and position means something different, and combinations can mean other things again, all affecting the degree and type of gene activity.

A small part of the Amphimedon queenslandica genome, with DNA positions on the X-axis, and annotation tracks in separate rows. Description and codes below.

One example of their data is show above. They tested the locations of five different histone modifications, (H3K4me3 and similar rows), as well as the RNA polymerase (RNAPII) that is the ultimate target of all the regulation. Also laid along the small segment of the genome (X-axis) are rows (tracks) that describe the predicted locations of genes (bottom rows, purple) and the expression of each location of the genome into RNA (dark and light orange, next rows up). At the very top, under the genomic coordinates, are predictions about chromatin states, based on the histone modification analysis, using knowledge from other (higher) animals. The code for this analysis is shown below
Codes for the top annotation tracks above.

The color codes for inferred chromatin states. "Active TSS" means an active transcription start site, or gene being expressed. "Transcr. at gene 5' and 3'" indicates signatures for boundaries of transcription units, though this is clearly not very precise. "Genic activated enhancers" mark upstream regulatory elements that regulate a local transcription start site, sometimes from far away. "Weak enahancers" mark the same, but less active. The "Bivalent/poised" notations indicate non-active transcription units or enhancers, which may have RNA polymerase parked, but not active. Lastly, the gray annotations indicate states of chromative repression of gene activity which are very important, and regulated by histone modification, but not relevant in this genomic location at the (adult) stage sampled.

A few sponge genes with differential expression in larval vs adult stages. Expression in orange and blue,  and the histone marks (H3K4me3) in purple at top. Genes are in purple at the bottom, complete with their complex exon-intron structure.

Another example of their data shows the developmental difference in a small genomic region between larva and adult stages. The triple methylation of Histone H3 at lysine 4 (H3K4me3) turns out to be the most clearly informative chromatin marker, and comes up reliably at active gene start sites/promoters, where it plays a role in activating transcription.

All this goes to show that the tools and knowledge that have been accumulated over the last couple of decades for the study of model animals like mice, fruit flies, and humans, are also totally relevant for the study of sponges at the base of the animal tree. And in turn, that sponges already had and have complex molecular mechanisms that drive their developmental and morphological complexity, such as it is.

If this study revolved around the histone code alone, it would not have been very novel, since yeast cells share much of this apparatus, and are tenaciously unicellular. But the researchers also delve into several other properties that are more diagnostic of animals, such as the the locations of enhancer elements. This sponge genome can have enhancers over 10,000 basepairs away from the sites they regulate, which allows multiple enhancer cassettes can operate on the same gene to provide complex combinatorial and developmentally variable control. Additionally, the large set of gene regulatory proteins is more reminiscent of the set available in animals, many of which themselves have complex enhancer-influenced control.

The large number of introns (interruptions in the blue coding areas in the diagrams above) is another sign of animal-like organization, and in the sponge turn out to be where many of the enhancers live that regulate genes at a distance. Lastly, the researchers mention that about 60 pairs of sponge genes are similarly paired in other animal species, (i.e., they are micro-syntenic), indicating another level of unexpected conservation over about 700 million years. These are some of the ingredients that molecular biologists are learning are important to climb the organizational ladder to multicellularity.

Saturday, May 13, 2017

Hanging On For Dear Life

DNA synthesis on both the leading and lagging strands is done by an integrated, quasi-stable complex.

At the very heart of our hereditary and developmental lives is DNA synthesis- replication that copies one duplex into two. But each beautiful, classic DNA duplex has two strands, each of which needs to be copied. Not only that, but those two strands go in opposite directions. The one is the complement, or anti-sense, of the other one, and in chemical terms heads the other way. This creates a significant problem, since the overall direction of replication can only go one way- the direction of the fork where the parental duplex splits into two.

Complementarity of DNA. When one strand is synthesized in one direction, the other strand has to be synthesized in the opposite direction.

Replication of one of the resulting strands is easy- the "leading strand" is synthesized continuously from 5' starting phosphate onwards as the replication fork progresses. But the other strand, likewise synthesizing from 5' to 3', must head in the opposite direction, away from the replication fork. It is thus called the "lagging strand". Not only must it go in the wrong direction, but it has to work in pieces, synthesizing a short piece away from the replication fork, and then start over again once more single stranded template has been uncovered as the replication fork progresses. These units of newly synthesized DNA on the lagging strand, which are short and separated before they are later repaired and ligated together, are called Okazaki fragments, for the first person to observe and understand them, Reiji Okazaki.

Lagging strand being made in fragments. A helicase unwinding the fork is shown in gold, and the polymerases are show as beige doughnuts. The lagging strand polymerase has to go backwards, away from the direction of fork progression.

Naturally, DNA synthesis happens best when done on a smooth continuous basis, with as few interruptions and errors as possible. So the complexes assembled at the replication fork are quite stable, and held on by special protein clamps which surround the DNA like little doughnuts. An interesting finding was that in addition to the leading strand DNA polymerase, the replication complex also contains the polymerase used for lagging strand DNA synthesis, even while they are heading in opposite directions! This obviously helps keep the whole process orderly, but is a little hard to envision, since in DNA terms, the two polymerases can be hundreds of base pairs apart as they chug along their separate tracks.

More detail on the replication complex. The two polymerases (yellow) doing leading and lagging strand synthesis are actually tethered into one master complex (green), and the lagging strand polymerase, once done with one Okazaki fragment, is yanked back to start the next from near the replication fork. The clamps (red) are loaded by the green central complex in advance of each lagging strand re-initiation.

A recent paper investigated the notorious stability of these complexes. If you dilute an operating replication complex to homeopathic concentrations, the proteins will still stay on track, and continue synthesizing DNA. That indicates that the proteins do not spontaneously fall off by diffusion during their work, or even during the intricate switch between initiation points that is required in lagging strand synthesis. On the other hand, if you add inactive enzyme to ongoing replication complexes, they will grind to a halt, indicating that some kind of polymerase enzyme exchange is going on. What gives?

These authors used a microscopic method and fluorescently labeled polymerases to look at this question in detail. They were able to use a flow cell to stage replication as a visual process, watching a single molecule of DNA extend over time downstream as it was synthesized. With fluorescent polymerase, the progression is even clearer, and a hangup was observed at the first and second Okazaki fragment boundaries (C, the middle lines), since they had not added the ligases that could resolve those boundaries and free up the polymerase to return to the replication fork for another round.

A single DNA molecule being replicated, over time, to longer lengths from a tethered end. When the polymerase is fluorescently labeled, (bottom, purple), it shows the edge of the fork (top) and also the Okazaki fragment boundaries (lower lines).

When a mixture of two different fluorescent polymerases is used, (green and purple), they seem to exchange on the replicating complex, as it goes forth.

But when they used a mixture of two polymerases with different fluorescent colors, the interesting thing was that the colors changed while underway. A fork that started with purple polymerase changed suddenly to green, and then later back to purple. This is very odd for a stable complex. And if they fired a laser at the complex, bleaching its fluorophore permanently, the fluorescence eventually returned, indicating that a new polymerase molecule had hopped on and replaced the original one.

Lastly, if the progressing polymerase is zapped with permanently bleaching light, other fluorescent molecules take its place relatively quickly as the replication proceeds. Bottom is just a graph of the fluorescence intensity from the upper single molecule image.

The replication fork is thus a little like a basketball team. Substitutions can be made while the action is going on, but there can never be less than a full complement of players on the floor. The complexes are stable, but only if there is no other polymerase around. If there are, a new polymerase can jump into the existing fork / complex. It is a clever design that allows resolution of stalled complexes and defunct enzymes, while insuring that the process of replication goes forth as relentlessly and stably as possible.


  • GOP vs not just medicaid, and exchanges, but medicare too.
  • And then, naturally, lies about it.
  • Please stand up for net neutrality.
  • Facts, Schmacts.
  • Puerto Rico in the hands of a very few.
  • Timeline of the fraudulent firing.
  • Our president: "James Comey better hope that there are no "tapes" of our conversations before he starts leaking to the press!"
  • Krugman on our new, great, GOP.
  • Inequality diminishes our society.

Saturday, May 6, 2017

Thinking in Symbols, Speaking in Tongues

Theology, schmeology. Jimmy Swaggart's musical, tribal, shamanistic approach to religion.

What is church without music? Probably not much fun. Even Islam has smuggled in a musical tradition in the form of the call to prayer, which is often a virtuoso vocal performance. The important role music has in most religions is a sign that their gatherings are social bonding events, not scientific conferences. One of the clearest instances is the televangelism of Jimmy Swaggart, still going strong after 45 years. At its core are his amazing piano and vocal abilities, combined with a very tight band and other featured singers. A cousin of Jerry Lee Lewis, his talent was recognized early on, but instead of joining the recording industry, he built his own evangelical empire, whose broadcasts are heavy on the music, and light on the theology. The music is an extremely comforting, a sort of bluesy, (this is from Louisiana, after all), cross between Lawrence Welk and African American Church gospel. The focus is on praise and succor from god ("Take my hand, precious lord", "Jesus, use me", "I'll never be lonely again", "Sheltered in the arms of God", "What a friend we have in Jesus").

It is a little reminiscent of the gatherings of DeadHeads, finding a comforting sacrament of friendship and love in an endless bluesy/country jam, heavy on the sentiment. Yes, the Pentecostal / Swaggart version is a lot more conservative, and its love doesn't come from a puff of smoke, but the equally vaporous triumvarate of holy spirit, Jesus, and god. What theology there is is virtually stripped of any sense, however, consisting of archetypal references to give the whole jam more emotional power. It is, essentially, the power of shamanism.


For example, the creed of the ministry is:
“Dear Lord Jesus, I now realize that I am a sinner. I accept the fact that You died for me on the rugged Cross of Calvary. I now open my heart’s door and receive You as Saviour and Lord of my life. Please take full control of me and help me to be the kind of Christian You want me to be. Amen.”

Why "rugged"? How can Jesus take control when he no longer exists, and we do not know where he is or what he wants? OK, call me skeptical! Anyhow, the answer is always prayer, and the Swaggarts claim "Without a daily communication with our heavenly Father, we will only go so far in this Christian life, which won’t be very far, spiritually speaking." Prayer is the cell phone call to god, keeping Him up to date with what we want, and telling us what He wants. What does this really mean? It means one's conscience is going to do the talking, (at best), and its quality is going to be the tenor of our supposed talk with god. This might expain the problems that Swaggart himself has had in the sin department. It also means that anyone who can infiltrate our conscience and purport to tell us what god wants may end up with a great deal of control over our actions. For example, the ministry offers a wide range of "The truth about..." videos, telling the flock why Mormons, Catholics, Muslims, even Seventh-Day Adventists, are wrong and bad. The "full control of me" formulation could be taken as a little sinister, not just to keep the tribal boundaries clear, but to milk the flock for money, and drive a highly conservative political message that is in many respects rather uncharitable.

And control is surely what is going on here. The bonding is very strong and social, starting with the music and the TV shows, but extending to more intense "Camp Meetings" and other events around the country. Advertisements for youth events gush about how attendees feel the presence of the lord during the event. Theoretically, god is supposed to be everywhere, not anywhwere in particular, so these charismatic settings and climaxes are a clear sign of shamanism, not of any coherent theology, let alone philosophy.

The magazine, aside from advertisements for a rich assortment of ministry products, is full of word-salad theology, with submission urged to the will of Christ, or other spirits, so that the "Blood of the Lamb" can wash over the sinners in the pews, cleansing them of their sins. Sermons and blessings are "anointed", messages are "Spirit-baptized", and the "Powers of Darkness" are fought. This church is the "Bride of Christ", and good believers have "eyes of the spirit" to see "the things that lurk in darkness". The whole thing is a work of art, really- a poetry of metaphor which is highly meaningful without actually meaning anything concrete or real in this world. But when it comes to the prices, things are naturally far more explicit "Your price just $10 each".

Televangelism remains a remarkable phenomenon, drawing on the implicit cultural assumptions in favor of Christianity, on blues and gospel music, and on the power of personal magnetism and group bonding to comfort the lonely and lost. If the message were stripped down to the hymns alone, it would be a positive social force.
"If God Is Dead Who's This Living In My Soul?"


  • Our government, regular people not invited.
  • Even the National Review wonders about inequality. And then concludes that it should be made worse.
  • GOP busy making things worse.. Things which clearly could be better.
  • Making Afghanistan great again.

Saturday, April 29, 2017

The Rich Get Richer

Inequality is immoral and unjust: Review of "The Anatomy of Inequality: Its Social and Economic Origins- and Solutions", by Per Molander

Second only to global warming and our vast debt to the future of the biosphere, inequality is the most pressing issue of the day. The two issues incidentally interact, since the slide towards inequality has correlated with a glaring lack of care and empathy by the upper classes, not only towards their fellow citizens and humans, but towards all other life on the planet.

This book is the third of a powerful triumvarate of recent analyses of economic inequality, and of them perhaps the most profound. Its point is very simple- that inequality is absolutely intrinsic to our socio-economic relations, and proceeds naturally to extreme levels unless stopped by special political or social interventions. Inequality is not fair, it is not reasonable, just, or virtuous, but rather is a mechanistic and inexorable consequence of game theory, by which, as the title and the saying have it, the rich get richer.

Life is full of negotiations- with family members, with sellers of products, with employers. Not all negotiations are obvious, but when we are faced with a difficult choice, there is probably some kind of negotiation at the heart of it, influenced by other people. A paradigmatic negotiation is for a job. A seeker has labor to offer, and an employer has money to offer in return. What is the employer's need for labor vs the seeker's need for money? Already one can see a deep asymmetry at work. Add selective secrecy that hides the pay of other employees, an entire department of people devoted to getting the best deal for the company, and a social hierarchy based on seniority, and the result is clear- a lot of underpaid employees.

In negotiations, the more powerful / richer party is substantially advantaged. The employer may not only have the advantages above, but also have political influence to engineer other non-market advantages, from H1B visas to union-busting legislation and regulation (or non-regulation). This process was far more stark historically, obviously, as kings and other nobles gave themselves ownership of all the land, and made peasants into serfs. The advantages of a strong pre-existing negotiating position are so persistant that, in the natural state of affairs, inequality inexorably goes towards infinity, with the only limit being the dominant party's interest in keeping the whole iniquitous system going. As Molander puts it:
"There is no reason to believe that people in hunter-gatherer cultures have a different constitution or capacity to exert moral pressure that would curb the power of a ruling class. These societies are egalitarian quite simply because there is very little room for inequality when a society is close to the subsitance level. In societies with a larger suplus of goods, what restricts the power imbalance between the affluent and the impoverished is the former's interest in keeping the latter alive and reasonably fit for work."

Herein lies the two-income trap, the stagnation of wages over the last few decades, and the increasing share of wealth and income going to capital in the US. It is a mechanism analogous to natural selection, both of which amplify the smallest differences (or vagaries of fate and luck) into existential propositions. In any negotiation about economic issues, the person with a bigger cushion of money is better off, even when all other aspects are identical. Who needs the money more and will give in first? Clearly the person with less of it. Who is better able to take investment risks, and get higher long-term returns, even at the expense of volatility? Again, the one with more money. Having money, whatever its source, is like having the wind at your back, giving rise in turn to more money, higher status, more opportunities, and better bargaining positions vs others, and thus widening inequality.

While we in this new gilded age are nowhere near medieval serfdom, we hardly exist in a fair system, either. Take the example of Bill Gates. Smarter and more productive than your average person? Yes. But 80 billion times more? Probably not. His (borrowed) operating system was, even at the time, hardly the best available. But he parlayed a lucky industry paradigm and a strong but bumbling partner- IBM- into the world's biggest fortune.

Now, for all the money he is giving away, he can't help but make billions more each year from investments. Can you or I stand back and have billions come in the door? Probably not. Such passive income, derived not from investment risk (given the wealth backing him up, there is zero risk to practically any financial action) or from accumen (which can be purchased), is almost purely parasitic on the labors of everyone else in society. Add in the ability to bequeath this horde to his children, (think Paris Hilton, and the Republican obsession to eliminate estate taxes entirely), or to use it to derange our political and social system (think Kochs, or Mercers) and the corrosive nature of this concentrated and unearned wealth is obvious.

Historically, there have been several solutions to this ratchet of inquality. Catastrophes, such as plagues and wars, have been great levelers. Our own Depression, World War, and Cold War era was when many large fortunes dissolved and the stage set for several decades of low inequality and general prosperity. But in normal times, the ratchet of wealth accumulation runs inexorably and unfairly. Molander mentions the Jewish tradition of a debt Jubilee, which reset land ownership and debt bondage in Israel every 50 years. Outright revolution is another method, such as experienced in France and Russia, or the various democratic / demagogic episodes in Greek history. Or on a more modest scale, land reform, which shares out what was previously held by large landowners.


Molander notes that constitutions and social constracts are of limited help against this natural process of differentiation, which in his engineering terms is a positive feedback. We need more dynamic, negative feedback social mechanisms to resolve this fundamental economic and social injustice. The end of his book is taken up with peans to the social democracy of his native Scandinavia. But one problem is that social cohesiveness is both a product of, and a precondition for, substantial economic equality. We can not wait around for the number of social democrats in our national instutions to exceed one. As previously discussed, we in the US used to have, around our founding, the conditions of rough equality, which led in large part to the enlightened consitution and unwritten institutions we cherish and have benifitted by so much.

But now we don't, and our conditions, barring some unwelcome catastrophe, are trending strongly in the wrong direction. We need a new institutional structure that fights systematically against the natural ratchet of unequality. This would be something like the Federal Reserve, which has numerical targets that it is structured to hit in its management of the macroeconomy. For the Fed, this is the interest rate, of 2%, plus other goals of financial stability and (a very distant third) high employment. The Department of Economic Justice would target a basket of goals, including the Gini index (about 0.3), and employment (an unemplyment rate of nor more than 4%, or 10% in broad measures), and wealth distribution (no quintile owning more than the twice the next-lower quintile).

There are many tools that can be used to hit these targets. But it is significant at the outset that explicit, ambitious, and sustaintable targets be set. Some inequality is a good thing, fostering human development, work, and innovation. But too much degrades economic prosperity, and far more importantly, our social and political environment. The Department of Economic Justice may not have direct powers, perhaps only to report on conditions and propose legislation and regulation. Or it may have regulatory powers, such as for antitrust and fee or tax-based regulation of the financial industry. There are many available tools, from the obvious like taxing high incomes and taxing wealth directly, to financial transaction taxes, raising minimum wages, dramatically increasing public goods and infrastructure, and providing a job guarantee, which help raise the floor of economic conditions.

Molander provides a crucial service in reframing the inequality phenomenon as one that is natural and inexorable, but also mindless and unjust. The Galts of the world have very rarely invented the steel on which their fortunes rest, let alone the fortunes which their fortunes in turn make on a passive basis. We do not have to collectively put up with it.


  • Object lesson in the brutality of power inequality, in even small amounts.
  • Workers are underpaid.. what else is news?
  • Some great signs.
  • Honestly, are we capable of self-government?
  • News from Afghanistan, cont.: corruption rampant.
  • Reflections on truth.
  • Stiglitz on a modern welfare (or, perhaps, fair) state.
  • The last mask is off. The tax give-aways go to the rich and to Trump, not to his voters. But maybe they don't care.
  • Which makes Trump "perfect".

Saturday, April 22, 2017

How Speech Gets Computed, Maybe

Speech is just sounds, so why does it not sound like noise?

Even wonder how we learn a language? How we make the transition from hearing how a totally foreign language sounds- like gibberish- to fluent understanding of the ideas flowing in, without even noticing their sound characteristics, or at best, appreciating them as poetry or song- a curious mixture of meaning and sound-play? Something happens in our brains, but what? A recent paper discusses the computational activities that are going on under the hood.

Speech arrives as a continuous stream of various sounds, which we can split up into discrete units (phonemes), which we then have to reconstruct as meaning-units, and detect their long-range interactions with other units such as the word, sentence, and paragraph. That is, their grammar and higher-level meanings. Obviously, in light of the difficulty of getting machines to do this well, it is a challenging task. And while we have lots of neurons to throw at the problem, each one is very slow- a mere 20 milliseconds at best, compared to about well under a nanosecond for current computers, about 100 million times faster. It is not a very promising situation.

An elementary neural network, used by the authors. Higher levels operate more slowly, capturing broader meanings.

The authors focus on how pieces of the problem are separated, recombined, and solved, in the context of a stream of stimulus facing a neural network like the brain. They realize that analogical thinking, where we routinely schematize concepts and remember them in relational ways that link between sub-concepts, super-concepts, similar concepts, coordinated experiences, etc. may form a deep precursor from non-language thought that enabled the rapid evolution of language decoding and perception.

One aspect of their solution is that the information of the incoming stream is used, but not rapidly discarded as one would in some computational approaches. Recognizing a phrase within the stream of speech is a big accomplishment, but the next word may alter its interpretation fundamentally, and require access to its component parts for that reinterpretation. So bits of the meaning hierarchy (words, phrases) are identified as they come in, but must also be kept, piece-wise, in memory for further Bayesian consideration and reconstruction. This is easy enough to say, but how would it be implemented?

From the paper, it is hard to tell, actually. The details are hidden in the program they use and in prior work. They use a neural network of several layers, originally devised to detect relational logic from unstructured inputs. The idea was to use rapid classifications and hierarchies from incoming data (specific propositional statements or facts) to set up analogies, which enable drawing very general relations among concepts/ideas. They argue that this is a good model for how our brains work. The kicker is that the same network is quite effective in understanding speech, relating (binding) nearby (in time) meaning units together even while it also holds them distinct and generates higher-level logic from them. It even shows oscillations that match quite closely those seen in the active auditory cortex, which is known to entrain its oscillations to speech patterns. High activity at the 2 and 4 Hz bands seem to relate to the pace of speech.

"The basic idea is to encode the elements that are bound in lower layers of a hierarchy directly from the sequential input and then use slower dynamics to accumulate evidence for relations at higher levels of the hierarchy. This necessarily entails a memory of the ordinal relationships that, computationally, requires higher-level representations to integrate or bind lower-level representations over time—with more protracted activity. This temporal binding mandates an asynchrony of representation between hierarchical levels of representation in order to maintain distinct, separable representations despite binding."

This result and the surrounding work, cloudy though they are, also forms an evolutionary argument, that speech recognition, being computationally very similar to other forms of analogical / relational / hierarchical thinking, may have arisen rather easily from pre-existing capabilities. Neural networks are all the rage now, with Google among others drawing on them for phenomenal advances in speech and image recognition. So there seems to be a convergence from the technology and research sides to say that this principle of computation, so different from the silicon-based sequential and procedural processing paradigm, holds tremendous promise for understanding our brains as well as exceeding them.


Sunday, April 16, 2017

Are You a Turtle?

The many types of personality types. Review of "The Five Elements", by Dondi Dahlin.

It may be trite to say that people are different. Yet it bears repeating, as we routinely lose sight of that simple fact in our private echo chambers of righteous self-regard. We routinely assume that everyone is on some level the same as we are, and then are left befuddled and shocked when an election turns that idea on its head. In our personal lives, we routinely assume that others see what we see, and care about what we care about, and get angry when they blithely ignore that sock on the floor, or the monthly budget. Thus it is perennially important to raise consciousness about the depth and durability of human differences, so that we can live together more harmoniously, with greater understanding.

Personality and temperament have been an ongoing theme on this site, so it was a pleasure to find a new book on the subject that is as compelling as anything I have read. There are many systems that try to schematize the protean landscape of human personality, from the ancient (the four humors) to the pseudo-scientific (Jung's) and more systematic (Myers-Briggs), and the more frankly scientific (the five factor model). Dahlin is a California New-Age scion whose family has long used the Chinese five-element system to understand themselves and others.

Who are you? Who am I?

There are virtues to using such a simple system. On the surface, it captures extreme personality types, which can then be combined to come up with a more nuanced model of a particular person's style. To recount them very briefly, the Water type is introverted, slow, inner-directed, and creative. The Wood type is type-A, always doing, with a task list in hand, and impatient with ritual. The Fire type is the happy-go-lucky, life-of-the-party type, intensely interested and charming one minute, then off to something new the next. The Earth type is the home and family-oriented type, the human doormat, always serving others. And the Metal type is cool, idealistic, minimal. They subscribe to Dwell magazine.

There is a great deal more to say, of course, and Dahlin says it very well, with warmth, verve, and fascinating anecdotes. Each type has its spiritual approaches, its physical and relationship needs, its manifestations in childhood. Dahlin even supplies special exercises for each type, drawn from her experience as a dancer. This book, more than the others of this genre that I have read, prompted me to think about other people in my life through its lens, which brings out particuliarities and deeper themes that tend to get lost in the hubbub of normal relations.

Are these personality types genetic? Dahlin's mother writes in the forward about the experience of being pregnant with children of different elements. Their personalities were markedly different well before birth. Dahlin herself notes, however, that personalities can change with time, during childhood and after trauma, which suggests that they are canalized styles of expression, solutions to the problems of life, which are not entirely determined. Indeed, the genetics of twins indicates that stable personality traits are determined only about half genetically, and half by other influences, whether environmental or stochastic.

Dahlin notes that most couples are of different types- opposites really do attract, and often complement each other as well. This suggests two things- that the insofar as personality type is genetic, the stable population distribution of different types is a matter of balanced selection. This should not come as a surprise, as many other species show personalities and stable population distributions of different types, down to ants and their castes, as an extreme form. The second implication is that appreciation of different types (using whatever system you like) is highly important at all levels- to our personal lives as well as our public and work lives, to promote understanding and lift a veil of mystery from our often unthinking view of the "other".


  • Fairness, inequality and just deserts.
  • We are still in a zero world, and need more fiscal support.
  • United uses a complex market to get passengers into its seats, but when it wants them out, it calls the police.
  • Where is the GOP going with Social Security? Not towards security.
  • Tax cuts serve the rich, unsurprisingly.
  • The disaster of privatization.
  • The disaster of deregulation. Who can call this populist?
  • We are going Byzantine.
  • Remember when Trump promised everyone even better health care?
  • An economist on carbon taxes.

Saturday, April 8, 2017

American Exceptionalism Rests on the Middle Class

What happens to our country when the middle class disappears? Review of Ganesh Sitaraman's "The Middle Class Constitution".

We are in a crisis right now in the US. Our government has been taken over by plutocrats who are busily reversing every progressive and public-spirited policy they can get their hands on. The worst nightmare of the nation's founders, revolving around demagogues, rule by extreme wealth, deepening political corruption, and mass immiseration, is coming to pass. Thomas Piketty's work on the nature and growth of economic inequality was the first major intellectual shot against this development, from an econometric perspective. Sitaraman's new book is the second, putting our economic inequality and its political consequences into historical and legal perspective, as a far greater matter than mere economics- rather, as a constitutional crisis.

The US has been since its founding exceptional on the world stage, not for our apple pie or iPhones, but for our democratic social system. Now the world is filled with democracies, so our position as one of the more mediocre ones in terms of governance, public services, health status, and overall happiness is perhaps not such a surprise. But at the beginning, we were truly revolutionary, and Lincoln was not far off when he posed the Civil War as a test whether government by the people would survive globally. (Switzerland, however, probably would have pressed on with its own experiments).

Why did this experiment happen here? According to Sitaraman, and to Toqueville, and to many of the founders themselves and other observers, it happened because the economic condition of Americans was substantially one of equality. Immigration put most people on a similar footing, which is to say, poor. The availability of land to all who wanted to work it provided a base of subsistence to these immigrants, and a foundation of modest wealth. The ideology of the early colonies and Republic was one of severe allergy to class distinctions, titles, and the like. There were some wealthy people, but nothing like the disparities that we would see later, either in the guilded age, or today.

Our constitution reflects these origins, and depends on continued rough equality with a predominant middle class. Sitaraman makes it clear that historically, constitutions have always reflected the economic conditions they sought to rule. Where feudalism and serfdom was the rule, so was autocracy. And oligarchy or autocracy have been the rule through the vast majority of human history, with democracy only possible at the lowest levels of rural and small-town society, where equality was likewise occasionally possible. Golden-age Greece, our democratic ideal, was more like an oligarchy of the free men of the city. Karl Marx had a point when he maintained that ownership of the means of production controls the nature of the class system, and thence the political system. Even Aristotle recognized, that while a true middle class polity was impossible in the power-economics setting of the ancient world, it would be the ideal polity, composed of that class of people who are neither so greedy and power-mad that they would factionalize and corrupt the system, nor so poor that their only wish would be to revolt and overthrow it.

Our rough equality continued through the frontier era, when anyone could pull up stakes and get land out west. Conditions were increasingly difficult the farther west one went, until the coastal paradise of California and Oregon gave one last great frontier of fertile land. This availability of land amounted to a job guarantee for its day, allowing any willing person to work and make money regardless of the willingness of an employer to hire. Farming certainly had its own terrible risks, between the weather, pests, and fluctuating markets. But no able person had to be destitute.

Now, we are neither an agricultural society, nor have any frontiers left. We exist in a new form of feudalism where employers have no governance responsibility to their employees. Employees are purely at will, and may be fired at any time for any reason. It is ironic that we so fetishize political freedom and legal equality, while practicing, in our corporate culture, the most retrograde feudalism. For all the regulations that modestly ameliorate this condition, most workers are in a dramatically assymetric and powerless position. It wouldn't be so bad if the economic system were not also unstable, experiencing crises of demand and investment that force millions of workers into destitution. Additionally, the closing of the frontier and other restrictions have made land, housing, and rents increasingly subject to scarcity costs far above their construction costs. This makes destitution a far more probable and chilling prospect.



Sitaraman cites disturbing research into our political system that makes the case that we are already living in an oligarchy/plutocracy. The fact is that each of our politicians is for sale, the media system is run by corporations, and much of the governmental regulatory apparatus that was supposed to protect us from the predatory special interests have instead been captured by them. This latest research shows that the actions of our political system accord essentially none of the time with the preferences of the lower 90% of the population, and all of the time with the preferences of the rich and super-rich. Our so-called leaders, thanks to various forms of legal corruption, are steeped in the social melieu of the rich, its super-PACs, and its propaganda organs, from cable channels, to think tanks, to lobbyists. The rest of us hardly stand a chance. Even at state and local levels, the rich have been funding dramatic new advances in corruption and propaganda, which have turned our nation into a sea of red.

This is how wealth translates to power, and the US may have entered a terminal loop where the plutocracy is so entrenched and so shameless about leveraging that power into yet more power, that there is nothing further to do, short of revolution. The breathtaking nepotism, incompetence, greed, and immorality of our current administration merely puts an exclamation point on a decades-long process that has not only reshaped the political system into a frank plutocracy, but reshaped the economic system as well into one that freezes out the middle class, by dramatically lowering taxes on the rich and reducing public facilities and services, among many other policies.

Middle classes do not happen by accident. The natural course of events, given the Malthusian pressures documented so dryly by Thomas Piketty and many others, is towards competitive differentiation, with winners gathering more power and wealth, which, once it reaches a high level, grows by natural accretion and compounding (where it is not more actively leveraged) far beyond anyone's needs, and losers finding it ever more difficult to find a way into a brutally rigged system. Classically, this was expressed by ownership of land, to which the answer has been land reform, which is to say, expropriation. Sitaraman provides a fascinating aside on the sequel to the Civil War.
"The most eloquent advocate for confiscation and redistribution was Thaddeus Stevens. The clubfooted Pennsylvania congressman proposed confiscating the estates of the top 10% of wealthy rebel planters, which at the time amounted to those with more than $10,000 or more than two hundred acres of land. With that land, which he pointed out would leave 90 percent of southerners untouched, every freedman could be given forty acres. The remainder wold be sold at auction and used to fund veteran pensions, compensate the injured, and retire the war debt. Steven's reasoning acknowledge that this ction would be revolutionary, but he also deemed it necessary for preserving republican government. "The whole fabric of Southern society must be changed," he declared in a speech to constituents in 1865: 
"Without this, this government can never be, as it has never been, a true republic. Heretofore, it had more the features of aristocracy than of democracy. The Southern States have been despotisms, not governments of the people. It is impossible that any practical equality of rights can exist where a few thousand men monopolize the whole landed property. ... If the South is ever to be made a safe republic, let her lands be clutivated by the toil of the owners, or the free labor of intelligent citizens. This must be done even though it drive her nobility into exile. If they go, all the better."

The book ends with a relatively standard plea, in the Bernie Sanders/Elizabeth Warren frame of argument, for higher consciousness of these facts and trends, in order to frame a new phase of social activism. Sitaraman has great respect for the Progressive era of the late 19th century, which brought us the progressive income tax, anti-trust, and the regulatory state, among other innovations. Thinkers and movements of this time took a fundamental look at our system and recognized that one more agency would not be enough- we needed constitutional amendments and deep reform. Today, with the Supreme Court wielding the First Amendment like bludgeon against the very citizens it was designed to protect, drowning them in corporate doublespeak, and with our politico-economic power system declining into bannana republic levels of dysfunction and disparity, it is time again to crank up the volume of protest, and direct it to fundamental and radical aims, such as taking money out of politics, remaking corporate governance on a more democratic model, and restoring a tax and public financing policies that sustainably strengthen the middle class.


  • A clown is in charge of our economic policy.
  • Pay should not be a big secret.
  • Another lie ... and another.
  • We need a new privacy regime.
  • China is the last country to want any change in North Korea.
  • They really are better than everyone else.
  • Corrupt enrichment.
  • Health care is one of the bigger drivers of inequality.
  • Economic graph of the week. Corporations are saving more, growing fatter, while everyone else grows thinner.


Saturday, April 1, 2017

Con Men Bring Us Hope

Herbalife, Christianity, Scientology, and the Trump Network. Only I can connect the dots!

The New Yorker carried a remarkable article recently, ostensibly about Herbalife. That is the network or multilevel marketing company that makes its money luring would-be entrepreneurs to new careers selling a powdered food-related product, and charging them for every step towards their doom. The story was structured around a short that hedge fund star Bill Ackman attempted against Herbalife, thinking that a good dose of negative publicity could cause the pyramid scheme to collapse.

But the short ended in tears, as the stock recovered robustly and still trades at relatively high levels, with a respectable P/E of 19. It turned out that investors all knew the company was a scam. They didn't need to be pursuaded with flashy slides and conference calls. But getting the message to Herbalife's actual customers / marks was another matter, far beyond Ackman's PR capacity. The beauty of network marketing is that every person in the operation is motivated to spread the gospel. It is a theo-capitalist virus, made for growth on the back of desperate people's dreams.

Where were the regulators in all this? Well you may ask. The current Republican regime doesn't think much of any kind of regulation, representing, as they do, the predators of the system. The story brings up the extremely interesting tale of Donald Trump's own foray into network marketing, a window into his character like no other. This was an even more tenuous proposition, claiming to convey Trump's valuable business insights to its entrepreneur members, thereby allowing them to become rich in the wake of the 2008 recession. The scheme naturally ended up in court, then sold off to some other paragon of capitalism. What interest would such a person have in the victims of his form of capitalism? What credence should his voters have in his promises to help the little people?

Trump's character on display- always looking for a sucker.

Anyhow, the FTC did finally look into Herbalife, and issued a fine as well as directives to clean up its business. The fine was apparently sent to Herbalife distributors (who are also customers) who have been using it to buy more Herbalife products. The re-organization forced Herbalife to reclassify thousands of its distributors as customers so that it would have some customers actually buying products, instead of a multilevel network of distributors all selling down-line to suckers who thought they were also running a business. But naturally Herbalife found a way around this as well, spinning the whole thing as an FTC endorsement of their business model, and happy to putatively do the re-organization. Shares went up and all is well.
Herbalife is robust to whistleblowers, truth-tellers, the FTC, short sellers, and bad publicity in general.

The lesson from all this is that con men are virtually ineradicable, especially when we have toothless regulators. People are naturally, and necessarily, optimistic. America runs on optimism, yet optimism breeds the deceit and lies that are so easily passed off by psychopaths like Mr. Trump and the many other bottom-feeders of our competitive, capitalist system. Without regulation, a Gresham dynamic takes hold, honest business people can't make a dollar, and the entire system descends into a mire of curruption. Just look at the business climate in a country like Russia or Afghanistan.

But what if there are no regulators, because the con is not (ostensibly) business-related? What if a religion is being sold, with the most absurd visions of heavenly bliss (if you believe) and personal empowerment? Scientology started out as a sort of new-age psychotherapy. But when it met with derision from the from professional psychological community, and official sanction and bankruptcy for offering medicine (and selling E-meters) without a license, it was re-organized as a religion. Presto- not only could it continue as a quack medical cult, but it became tax-exempt as well! Now its members can believe in their hidden thetan abilities, purified "clear" states of consciousness, and extraterrestrial origins without skeptical questions from official bodies.

Religious cons are unimaginably more ornate and bold than business cons. Just think of the Jesus story. Who could have come up with such a thing? But given enough time, and enough idle dreaming and theological imagination, and, well, anything is possible. Now the entire world is saved, except for the non-believers who are all going to hell. We are so crushed by reality, so filled with hope that life should be better, and can be better if we only dream hard enough, that those who have crossed the line between reality and imagination have only to assert their confidence, for someone to believe them.

So the con men, mystics, and preachers, whether cynical, self-deluded, or psychopathic, make a matched set with the credulous, lost, and desperate. And aren't we all desperate, at some level? Desperate that this life, filled as it is with trials and pain, should be so difficult and end so definitively? What can the truth say about that? It is a perennial question, whether we put ourselves into the hands of yet another purveyor of hope, or practice moderation in this, as in so many other things.


Saturday, March 25, 2017

Centrioles Have Mothers Too

Penetrating the mysteries of one of our most attractive organelles.

The centriole is one of the more glamorous, yet enigmatic, structures of the eukaryotic cell. Yes, it has beautiful ultrastructure. Yes, it serves as the fabulous astral center of the spindle poles that organize mitosis. And yes, it has been studied for well over a century. But does all that mean it is well understood? Absolutely not.

Electron micrographs of centrioles. The inset is a cross-section of one centriole barrel, showing the 9 X 3 microtubule structure as well as the inner cartwheel that seems to be its construction scaffold. The main image shows a side-ways cross-section of a mother-daughter pair of centrioles, (forming a centrosome), showing their length and relationship. Also shown is the nimbus of microtubules coming out of what is called the peri-centriolar material, or PCM. How those relate functionally to the core structure is still not known.

Centrioles are one of the many radical innovations of eukaryotic cells, forming the core(s) of centrosomes, which organize mitosis, and of cilia, which are the novel eukaryotic solution to cellular locomotion. They are are a barrel-shaped quasi-crystalline arrangement of microtubule proteins, grown on a central wheel of template proteins, called the cartwheel. For cilia, these microtubules clearly serve as the foundation stones of microtubule rods that grow out into the long appendage. But things get far more murky at the mitotic spindle core, where a profusion of microtubules emerge from an amorphous zone around the centrioles, but are not templated directly by them. The nature and history of this structure remains quite mysterious, especially as it turns out that some eukaryotic cells can get by without centrioles at all.

Centrosomes are in red, at what are also called the microtubule organizing centers, organizing the green microtubules that  in turn organize the separation of DNA (blue) during mitosis. 

In addition to these starring, if not obligatory, roles, centrioles have another charismatic property, which is that they reproduce in a clearly parental fashion. Every cell cycle is accompanied by the division of the centriole pair into single centriole mothers, which then give birth to new centriole daughters from their sides. While most of the molecular actors in this centriole mini-drama are known, as are some of their roles, a great deal remains to be found out about how the whole process is put together.

Detailed scheme of centriole (green) and centrosome (the whole green + yellow mess) duplication during the cell cycle. The structures are reasonably well known, the molecules and their roles less so. Note the roles of Plk4, Sas6 and colleagues in generation of the nascent cartwheel scaffold at G1/S. Plk4 itself will be disposed of later on, towards M phase.  

A recent paper described some more about how the first steps of this replication process take place. The cooperation of the centrioles with the larger cell cycle is naturally very deep. Some labs have found that the key kinases (which attach regulatory phosphates to other proteins) that regulate the cell cycle operate out of the centrosome- the structure containing the centriole pair. And the centrioles in turn are subject to key kinase steps that license when they can replicate. Key molecules of this process have the name Sas, after a genetic screen for spindle assembly defective mutants, each of which was given a Sas# name. Sas4, Sas5, and Sas6, for instance, are proteins that make up most of the central cartwheel scaffold upon which the centrosomal microtubules are built, along with another protein, Cep135. But the story starts much earlier. Sas6 is brought to the key location on the mother centriole's side by a kinase, Zyg1, which in turn is brought there by Spd2. How did Spd2 get there, and what does it do? Could it be turtles all the way down?

No, the authors identify Sas7 as the protein that binds to Spd2 and gets centriole pregnancy underway. Sas7, finally, is at the centriole all the time, and is activated not by recruitment, but by the cell cycle. One key finding was that all the other known initiating mutations (of Zyg1 and Spd2) depend on Sas7 for their action and localization, but not the reverse. They also find that the centrioles of mutant cells are significantly diminished, missing quite a bit of the outside structure. This suggests that Sas7 is a structural component at just the right position, around the outsides of the mother centriole, to participate in the construction of daughter centrioles.

Complete deletion of Sas7 renders the organism dead. But a partially inactivating mutation (temperature sensitive) allows its function to be observed. Here, the wild-type cells show flamboyant microtubule (green) organization during mitosis. The mutant at the restrictive temperature shows a mess, where centriole duplication has failed and the DNA (orange) is dispersed around one pole instead of being nicely pulled between two poles.

Mutant Sas7 also causes structural problems for centrioles. The outsides of the lower centrioles are severely depleted of material, whatever it is.

But that leaves one last question- what licenses / initiates the beginning of centriole duplcation, given this need for Sas7 and Spd2 interaction and given the need for tight coupling with the cell cycle, and why does it happen at only one position on the side of the mother, rather than all over? This article does not touch on those issues, and one has to revert to other reviews to gain some insight. Zyg1 is known in other species as Plk4, and seems to be the critical link to the cell cycle. Whether its activation is driven in particular ways is not yet known, but its destruction is known to be driven by the SCF complex that funnels many critical cell cycle proteins into the proteasomal trash bin at the the transition between G2 and M phase (with partner betaTrCP). However, how the localization is restricted to one position on the mother is not known at all. The only relevant fact is that supplying an excess of Plk4 can prompt initiation of multiple daughters. Thus one reviewer is reduced to speculating about a concentration-sensitive positive feedback mechanism that forces all the activity to localize in one place.

So, even after all these years, of both macroscopic and molecular study, this beauty still holds quite a bit of mystery. Will resolving it make us happier?