Saturday, July 26, 2014

Continuing Human Evolution: Fallacies and Prospects

Where are we going, and where do we want to go, genetically speaking?

Of all the politically correct stances, the genetic unimprovability of humanity is perhaps the most inviolable. Even while we work daily, even feverishly, to improve other aspects of our material and cultural existence, our biology remains an ethical red zone, where nothing can be done and no infringement placed on individual replication. As a response to the abuse of eugenics in the 20th century, and to the deep philosophical problems involved, this is understandable. But it does not do justice to the underlying science.

Two recent books have raised the issue again, as does our growing knowledge of biology generally. Marlene Zuk, in her book "Paleofantasy: what evolution really tells us about sex, diet, and how we live" explains that we have been evolving all along, up to the present day. (Review.) To think that we are no more than rock-throwing homonids trapped in cubicles and talking to our computers is a bit inaccurate. We have many issues related to insufficient evolutionary adaptation, such as back problems, diet issues with the modern fare of fats, salts, and sweets. But we have also adapted successfully to much else on  a recent time scale, like high altitude, diets of milk, and malaria. Perhaps also to reading, sewing, and music. But this only scratches the surface. There are very likely many other, more complex traits that have been honed and reshaped over the millennia of recent human evolution. The technology to detect all this in our genomes is only slowly arriving.

The implications of this are clear enough. Humans have been evolving and adapting to many conditions up to the present day, and doing so differently in different areas of the world, leading to the evident differences among human beings. Humans are not created equal, and the question arises whether differences extend to psychological and behavioral traits as well.

The other book is "A Farewell to Alms", by Gregory Clark. He puts a very provocative hypothesis about the success of English society over the last millenium. England ran roughshod over the rest of the world during its imperial heyday, authored the industrial revolution, and basically did away with the Malthusian conundrum of human reproduction continually outstripping food production. What happened? Clark conjectures a genetic hypothesis, that over hundreds of years, the rich in England consistently had more children than the poor, creating a constant diffusion of behavioral traits of competence downward through the social hierarchy.

As the linked review makes clear, Clark is an economist, not a geneticist. The genetics of such an hypothesis are entirely unknown. All we have to go on are the many observations of family traits and twin studies, which indicate strongly that many psychological / behavioral traits are heritable to a large degree. Once one accepts that, one can easily see that, whatever one's view of the "competence" and other bourgeois values of the rich and successful of England, humans all over the world have been evolving along some kind of path in these dimensions, continuously. Human nature is not set or unalterable, just as it is highly diverse. The two go hand in hand, as diversity + time + unequal reproduction = evolution. (A third book by Thomas Suddendorf, on what separates us from apes in the biological sense, is relevant as well.)

For example, not all students gain equally from an education; not all people learn music equally well, or play sports equally well. Bach, Beethoven, and Mozart each came from families of professional musicians. We are not created equal, other than in self-granted political and spiritual terms. The long fixation of the left on creating social progress solely through social policy and equalization, however well-intentioned, is not a fully realistic program, and in the long term, can have some dark consequences. On the other hand, the fixation of the right on making our economic competition as brutal as possible is not only cruel, but in genetic terms largely pointless, since in the moderately compassionate developed world, economic success is untethered from, indeed inverse to, reproductive success in broad terms.

Every society has a status system by which some behaviors or aspects are honored, while others are dishonored, to the point of death. Clark describes a society that, in his view, aligned its status system with reproductive success, thus building whatever those status-valued traits were into future populations. Assortive mating among high-status individuals created fierce competition for status, as did the consequences of failure: misery in addition to no reproduction. I am reading the tales of King Arthur currently, and that society fits the same template. Economic success was the direct fruit of political success, especially favor of the king, and was closely tied to the winning of damsels, estates, and reproductive success.


We do not take our own status system quite so seriously, as higher status individuals tend to reproduce less than those of lower status. One problem with Clark's thesis is that what he found in England is very likely the rule in all societies of that stage and earlier, where the rich lord it over the poor in countless ways, including those having to do with reproduction. So it is quite difficult to use his hypothesis to explain the particular and unique competitive strengths of the British empire, though it can help explain in very broad terms the rise and fall of hierarchical cultures generally. He has to add the additional hypothesis that the nature of status and enrichment in medieval England was uniquely selective for competence of some imperial nature rather than, say, corruption, martial ruthlessness, or courtly obseqiousness. That is a significantly more difficult case to make.

The genetic hypothesis may be thought to compete with another one, outlined by Jared Diamond in "Guns, Germs, and Steel". This posits that the great density, fertility, and species richness of Asia over the other continents like Africa and the Americas gave its humans special advantages that sped the development of agriculture, urban cultures, state formation, metal technology, as well as the epidemic diseases that overwhelmed proto-colonial subjects / competitiors. It was happenstance, in short, not genetics. I think it is fair to say that both hypotheses can easily coexist, since the genetic hypothesis of human (eugenic) evolution through linkage of status to reproduction does not necessarily say anything about races or colonial competitions. It is a process that has been universal over all pre-modern cultures, operating more or less in parallel everywhere. Every society has its social and reproductive hierarchy, which succeeding generations embody in genetic terms.

One can also note that eugenic policies remain quite commonplace in the world today, principally in the form of religious competition to reproduce. Catholicism tries to extend its faith by high reproductive rates, bans on birth control, etc., as does Mormonism. Islam allows the socially destructive practice of polygamy, allowing high-status males many more wives and children than low status males. They clearly take their status system very seriously. In China, in contrast, the state has enforced a one-child policy, now somewhat loosened, which has kept the status quo, genetically speaking, though the advent of Yao Ming out of the Chinese basketball program is a good example of assortive mating still at work. And Latin America is notoriously, if inadvertantly, staging a "reconquista" of the US, through immigration and high birth rates.

So evolution is happening, and the future of humanity will look different from its past. The question is whether religious & accidental eugenics are the only acceptable kind, or whether other forms are worth contemplating. Before going into details, I'll note that the urgency of this issue is extremely low. Climate heating is a far, far more urgent threat to our future happiness. The whole issue may also be moot in the face of technological development. Genetic engineering may eventually allow detailed and insightful reprogramming of our genomes, with far greater practical and ethical implications than tinkering with reproduction policies. And after the robots take over or we upload ourselves to the cloud, well ... then the biological evolution of humanity will be a quaint memory.

The degree of effort needed for a conscious reproduction policy is also very small. Selection pressures in the single digits or below can have strong effects over long periods of time. The point of such a policy would simply be to link whatever we deem valuable about ourselves as humans to our future state.

Should we take our ambient status system more seriously? I don't think so- quite the opposite, in fact. The psychopathy of the most successful people in finance and business is legendary. Money is simply not a good metric in this complicated age, of human worth and ability to create a happy, moral, and prosperous future for humanity. Education is a bit more systematic and fair as a criterion, but also tends to reinforce inherited status as much as reveal individual merit. What other criterion would serve, and what mechanism could be used?

Ethically and philosophically as well, it is extremely difficult to come up with any criterion that gives the state the power to say that one person is of greater existential value than another, however much we do so day in and day out in our economic, social, military, penal, and sexual lives. Additionally, lacking the implicit competitive mechanisms of economic, political and social success linking to reproduction, the state would need to step in explicitly. There is, however, an innate contradiction of the democratic state, founded on legal and existential equality, getting into the explicit business of judging and making us unequal.

Yet ... I met a lady recently who lives in her car, and who told me she has five children, none of whom can take care of her, being as impulsive and improvident as she, including several cases of drug abuse. It is heartbreaking, and a little disturbing. A policy that nudges us communally in a more positive direction would be beneficial in the long run.

I think the key issue is the vastly different perceived costs of having children between high and low status groups in developed countries. At the high end, each child implies hundreds of thousands of dollars in education, college, sports, enrichment, tech gadgets, etc. etc. On the low end, the expense of an extra mouth to feed is negligible, and may even be covered by the government. When expectations are low, so are costs. One solution is to create a more socially supportive environment by making education free, subsidizing child care, and making medical care free. Indeed, I would abolish private schools and make all schooling on the same public level; government run at the pre-college level, and non-profit in higher levels. This would at least level the playing field in family (cost) planning.

A second policy would be to guarantee work to everyone. Much teen pregnancy is the result of aimlessness and hopelessness, which could be ameliorated by integrating young people into a universal expectation of work and usefulness. Everyone should be trained if not college educated, and the government should offer jobs, on the pattern of the Depression-era jobs programs, to anyone not competitive enough to be recruited by the private labor market, but not disabled. Work at decent pay (substantially above welfare levels) should be a right, completing the safety net in a way that encourages responsibility and long-term life planning.

These two types of social policy would help relieve the frightening cost of having children for those with middle-class expectations, while encouraging everyone to engage in family planning and personal development that might go some distance to leveling the fertility playing field.



Saturday, July 19, 2014

Quantum Consciousness?

The demarcation problem in science meets New Age tech talk.

What is science and what is not science? The difference is not terribly clear, an issue called the demarcation problem. Is theology a science of the supernatural realm? Is psi research on extrasensory perception science, as it uses scientific methods? Is string theory in physics science, even though its chances of empirical validation seem rather slim?

It isn't very clear. Science tends to be whatever scientists do and view as valid in their expert communities. Whacky ideas may migrate in from fringe areas, (atoms, endosymbiosis, plate tectonics, ulcer-causing bacteria), turning from non-science into science once evidence appears. Conversely, long-hallowed ideas within science may turn out to be complete rubbish, like space-ethers, geocentrism, and the medical humors. There is fringe - mainstream traffic, though it tends to be rather light these days, since mainstream scientists generally know what they are doing.

Unfortunately the fringe areas are enormous, populated by people highly motivated to push pet theories that tend to have some psychological motivator. Psi research is a good example, which responds to our hopeful magical thinking that somehow, some way, even though those darn materialists don't have a drop of imagination in their brains, humans can indeed sense the emotions of others far away, levitate objects, detect water through dousing rods, and see behind playing cards. At least a little, right?

The scientific fringe is part of a broader cultural miasma of misinformation, from Fox news to Herbalife to Koch political subversion to mundane political campaigns and commercial advertisements. We live in a flurry of BS coming at us from all directions, and typically, following the motivation and the funding source is a critical tool to gauge the truthiness of claims. Russia's shameless campaign of lying about Ukraine is perhaps the moment's most egregious and deadly example. So science is far from alone in living in a perilous epistemological swamp. It just tries to do a better job by way of disinterested institutions, public practices, empirical adjudication, and all the other standards that come under the so-called scientific method. Can we deploy such methods on interesting topics, or are they intrinsically confined to uninteresting ones?

The mother of all demarcation nightmares has been creationism. Otherwise known as creation science, or intelligent design. The motivation is obvious: support traditional intuitions (and some scriptural readings) to deny that humans are animals. Credentialed scientists have been deployed, glossy textbooks written, museums established, articles and books written, evidence cherry-picked, school boards subverted, all to push a theory that the scientific community dimissed many decades ago. But given enough science-y paraphernalia, they could make a decent case, at least in the popular media, that they were engaging in science. A spineless political system was reduced to mouthing the mantra that schools should "teach the controversy".

Thankfully that controversy has died down in recent years, and the professional community feels less threatened by cultural bulldozing. Nevertheless, the needle has hardly budged in the population at large, of which 42% believe in creationism outright, and 31% more believe that evolution was guided by god, which is pretty much the opposite of the whole point of evolutionary theory as currently understood. And abroad, the Islamic world is almost uniformly creationist. It is a testament to the strength of psychological intuitions and archetypes, as well as media echo chambers.

Some recent discussions have gotten me interested in another area of motivated science, which is quantum consciousness. Here, it is obvious that our intuition (and a great deal of theology) militates against a materialist view of the brain and mind. The mind-body "problem" has been perennial fodder for philosophy. Could our minds be the subjective product of nerve firings in our brains? No way! Despite the rather obvious empirical parameters that show just that, intuitionally-driven models have always looked elsewhere, invoking souls of various sorts, which typically have the added bonus of immortality, another intuited archetype. The latest version of this is the movement among philosophers to posit a cosmic consciousness (Nagel, Chalmers), which in hand-waving fashion hypothesizes that somehow, consciousness is a basic property of the cosmos, with Jain-ist particles of consciousness in every object, implying that such things as plants, and even rocks, may be conscious. It seems like a total surrender to obfuscation and mysticism, descending from the grandiose premise that, because they have been unable to figure out how it all works, no one else can either. Einstein may be able to get away with such foundational cosmic speculations, but even for him, it took more than handwaving about how no one could explain the speed of light.

One science-y form of this is quantum consciousness, where the mystery of consciousness is creatively linked to the very hard-science-y paradoxes of quantum mechanics to come up with .. something again quite vague, but the idea is that since quantum entanglement can allow instant communication of a sort at great distances, and perform outrageous computations, that this resolves those amazing capacities of our minds. Quantum mechanics has been drafted into numerous pseudoscience fields of this sort, actually.

The specific example of this field that is most advanced, in its quotient of science-y tech-talk and academic paraphernalia, is the Orch-OR theories propounded by Stuart Hameroff and Roger Penrose. Penrose is a Sir, and an eminent physicist and mathematician. Hameroff is a professor of anesthesiology and psychology at the University of Arizona. Their output of papers has been prodigious, and they host an annual conference on the topic, funded by Deepak Chopra's foundations among other interested parties. They are not charlatans, really, but I think they have totally lost the thread in this case. They present a magisterial review of their own theory in 2014.

Penrose starts off by laying the premise of their case- that due to Kurt Gödels' work, the human ability to be certain about things is mathematically impossible, which necessitates a non-conventional solution to consciousness.
"Critical of the viewpoint of ‘strong artificial intelligence’ (‘strong AI’), according to which all mental processes are entirely computational, both books [by Penrose] argued, by appealing to Gödel's theorem and other considerations, that certain aspects of human consciousness, such as understanding, must be beyond the scope of any computational system, i.e. ‘non-computable’. ...  The non-computable ingredient required for human consciousness and understanding, Penrose suggested, would have to lie in an area where our current physical theories are fundamentally incomplete, though of important relevance to the scales that are pertinent to the operation of our brains." 
"As shown by Gödel's theorem, Penrose described how the mental quality of ‘understanding’ cannot be encapsulated by any computational system and must derive from some ‘non-computable’ effect. Moreover, the neurocomputational approach to volition, where algorithmic computation completely determines all thought processes, appears to preclude any possibility for independent causal agency, or free will. Something else is needed. What non-computable factor may occur in the brain?"

Well, the fact is that humans are not that certain about things. Religions may be, but that is an emotional, not a formal, issue. We operate by Bayesian statistics, where new evidence alters our beliefs, which are always tentative and evolving as we gain experience, at least for those who are empirically engaged at all. We are not operating from a tight set of axioms, per the Gödellian system, which we transcend to understand novel or paradoxical truths in some inexplicable way (it only seems that way on LSD!). So this premise seems rather nonsensical, and the whole project starts off on a very sour note. Not only that, but the authors then go on to propose a solution (with quantum qubits migrating in microtubules) that, first, is physically impossible in the brain, and second, doesn't evade Gödel's theory anyhow, being just another form of computation. In Gödel's terms, we are very incomplete systems, whether quantum or not, but seem to get by despite that.

Hameroff's part is to focus on microtubules, which he has identified as the locus of consciousness by way of his studies of anesthesia. In mainstream science, microtubules are cytoskeletal structures, play a central role in orchestrating mitosis and cell shape, and serve as roads for the transport of cargo, which is particularly relevant in neurons, where the distance between the cell nucleus / body and its far projections can be measured in feet. These cells need constant traffic of cargoes over the microtubule network to maintain function.

His proposal is that general anesthetics work by destabilizing microtubules in the brain, or at least their quantum computations. This is itself, apart from its implications in quantum consciousness, a fringe hypothesis. Current thinking in the field is very focused on ion channels and neurotransmitter receptors as the targets, though it has been difficult to pin down the specifics. General anesthetics tend to be membrane-soluble, which leads to hypotheses about their having very broad effects on membranes (not a strong theory on its own anymore) or on proteins embedded in membranes which would naturally bind to hydrophobic chemicals as they do to membrane lipids. It doesn't help one fringe hypothesis to be dependent another one like this, for even if consciousness is not solved soon, the target of anesthesia is likely to be, by normal progress in the mainstream of neuroscience / molecular biology.

One mainstream review states:
"Anesthetics are pharmacological agents that target specific central nervous system receptors. Once they bind to their brain receptors, anesthetics modulate remote brain areas and end up interfering with global neuronal networks, leading to a controlled and reversible loss of consciousness."

It is worth noting that Penrose and Hameroff's review is extensively referenced, with citations to some work that shows, for instance, that microtubules can bind anesthetics. But this was done at such high concentrations, and found among so many other proteins that also bind, that it looks like clutching at straws. They even resort to a little bit of lying, towards the end where they enumerate predictions of their theory:
"Actions of psychoactive drugs, including antidepressants, involve neuronal microtubules. This [prediction] indeed appears to be the case. Fluoxitene (Prozac) acts through microtubules [167]; anesthetics also act through MTs [86]."

The anesthetic cited here is anthracene, which is more a poison and general chemical than an anesthetic. It is not used in medicine at all. There are plenty of chemicals that will knock out frogs (which were the subject here) without telling us much about anesthetics as a specific class. The Prozac reference is highly problematic as well, since Prozac is called an SSRI for a reason. It binds to and inhibits serotonin uptake pumps, and that is thought to be its primary mode of action. If it binds (at again, very high concentrations) to microtubules as cited, that would be a side-effect, not the primary mode of action. Additionally, if microtubule dynamics are altered to some degree by this drug, why do all the other SSRIs with different structures work? The only thing they have in common is their binding and inhibition of the serotonin transporter. This kind of highly selective, indeed misleading, citing is a big red flag, to add to the red flag of psychological motivation.

At the core of the vast enterprise is the propostion that somehow, gravitation, quantum mechanics, and microtubules hosting qubits impinge somehow on their host neurons help their computations escape the Gödellian trap ... and simultaneously constitute atoms of consciousness:
"The Orch-OR [orchestrated objective reduction] scheme adopts DP [Diósi–Penrose objective reduction, which is a version of a quantum gravity theory] as a physical proposal, but it goes further than this by attempting to relate this particular version of OR to the phenomenon of consciousness. Accordingly, the ‘choice’ involved in any quantum state-reduction process would be accompanied by a (miniscule) proto-element of experience, which we refer to as a moment of proto-consciousness, but we do not necessarily refer to this as actual consciousness for reasons to be described."

So a choice made by qubits in this scheme is instantaneous, solving the timing issues that makes free will impossible in a normal materialist theory. It also reflects the Copenhagen interpretation of quantum mechanics where an observer must be invoked to collapse (reduce) the wave function of quantum entities like electrons. The tiny observers apparently add up, in the end, to what we experience as consciousness.
"Consciousness results from discrete physical events; such events have always existed in the universe as non-cognitive, proto-conscious events, these acting as part of precise physical laws not yet fully understood. Biology evolved a mechanism to orchestrate such events and to couple them to neuronal activity, resulting in meaningful, cognitive, conscious moments and thence also to causal control of behavior. These events are proposed specifically to be moments of quantum state reduction (intrinsic quantum “self-measurement”)."

One problem, among very many, is that this sets up another mind-body conundrum. In the religious soul theories, the soul is immaterial, so it is hard to explain how it receives perceptions from the brain and injects its decisions back into that brain, which is at least acknowledged as the conduit for human behavior and sensation, if not its computational processor. Some interface is required, like the pineal gland in the system of Descarte. But in such an interface, how are physical atoms moved by immaterial, supernatural entities? There is no easy way to deal with this, other than waving it away with assertions of pan-soul-ism, where there is no localized interface, and the soul pervades everthing in some magical way.

With the microtubules, the authors claim that they might communicate with each other across the brain via gap junctions, which are small portals leading directly from one cell to another. But not only does normal nerve conduction show little effect from these junctions, indicating that they are typically not highly connected with other cells, but microtubules from one cell do not enter other cells through such junctions, (they stop at the border), so there really can't be a direct network. So the authors back up and say that the microtubules might affect their host nerve function, which then makes the whole theory nearly pointless, since a mere potentiation of normal nerve function gets us back into normal neurobiology and whatever that can accomplish in generating consciousness.
"The most logical strategic site for coherent microtubule Orch OR and consciousness is in post-synaptic dendrites and soma (in which microtubules are uniquely arrayed and stabilized) during integration phases in integrate-and-fire brain neurons. Synaptic inputs could ‘orchestrate’ tubulin states governed by quantum dipoles, leading to tubulin superposition in vast numbers of microtubules all involved quantum-coherently together in a large-scale quantum state, where entanglement and quantum computation takes place during integration. The termination, by OR, of this orchestrated quantum computation at the end of integration phases would select microtubule states which could then influence and regulate axonal firings, thus controlling conscious behavior."

One might also note in passing that the superposition of vast numbers of coherent entangled quantum entities in the brain is judged impossible by experts in the relevant fields. They have been laboring mightily to set up qubit computers in vacuums near absolute zero with handfuls of electrons. The idea that this could be done easily on a massive scale in the liquid, warm brain would cause some surprise and shock.

In the end, despite the intense New Age interest in this kind of speculation, and its extensive scholarly apparatus, it is at the far-out fringe of brain studies. At a regular neuroscience conference, Hameroff attends, but the issue of quantum consciousness is nowhere else in sight. A physicist comes with a stray poster that also invokes quantum computation, but the session devoted to mechanisms of consciousness is cleanly and clearly mainstream. They are not interested. In demarcation terms, Hameroff and colleagues have academic positions and publish their thoughts, but these are not fruitful thoughts, as they use heavily cherry-picked data for support, and sponsor no evident empirical progress in their program, which thus remains an edifice of rather wild speculation.

I am knowledgeable, but not an expert, and to me, it looks like a big snow job more than a serious scientific theory, from premises through the elaborate contents, to the conclusions. At its heart, there is a -magic happens here- kind of quality to the invocations of quantum effects that are supposed to solve non-problems like free will, or significant problems like subjective consciousness that are best left as single problems rather than compounded with significant mysteries from radically separate fields like quantum gravity.

There is also an unwillingness to recognize the great deal of mainstream work that undermines the theory. For instance, consciousness is quite well timed in its occurrence relative to other brain events like perception and willed action. There is no reason to demand instantaneous action / computation when it is well known that consciousness trails perceptions by hundreds of milliseconds, and also trails various types of reflex actions and even the opening phenomena of willed actions by similar amounts. It has a function of global integration and monitoring, rest assured. But intuition is, as usual, a poor guide to what is really going on.

Another issue is the localization of consciousness. Is your liver conscious? Are your toe nails? I don't think so, which speaks to the plausibility of cosmic consciousness theories implying the consciousness of rocks, plants, etc. Indeed, most processes in the brain are unconscious. Yet all neurons have microtubules in profusion, indeed all cells do, so theories connecting their cosmic capabilities with consciousness turn on their specific arrangement or augmentation, which ends up little better, indeed far worse, than mainstream theories about the arrangement, connectivity, and other properties of nerve cells whose relationship to thought is rather more plausible.

Saturday, July 12, 2014

Chaperones Step in to Increase Evolution's Promiscuity

Mutations are usually bad, but can be made less bad with some protein folding help.

Proteins are the central mechanics of life, catalyzing the reactions, lifting the loads, purifying the fluids, and expressing the genes. They are also the most frequent targets of damaging mutations in their encoding genes. Some of these mutations happen outside the amino acid coding regions, in regulatory areas, but the most dramatic ones are typically within the coding region, changing the amino acid sequence.

Such mutations can not only change the function of a protein, but can also change its stability- its ability to fold and stay folded. Random polypeptides typically do not fold in any coherent way, in contrast to biologically evolved proteins, so the ability to fold tightly is weakened by most mutations.

Here is where saviors come into the picture- the chaperones, which are proteins specialized to help other proteins fold. The most massive of these feature a large cave whose internal surface can switch, with consumption of ATP, between hydrophobic and hydrophilic surfaces. It is like sticking your protein inside a jiffy-pop dome that can alternately encourage unfolding and folding in a protected, isolated space, until the protein gets it all together. Remember that the insides of proteins tend to be hydrophobic, and the outsides tend to be hydrophilic, which helps to direct the folding path in normal proteins, in addition to detailed spatial and electrostatic relationships among the amino acids.

Structural aspects of one chaperone, GroES/GroEL, within which other proteins (purple & yellow, in d) other proteins can fold in isolation. As shown in C, the interior can switch, using ATP, between hydrophobic and hydrophilic surfaces, encouraging folding and unfolding, respectively. Thus even without knowing how the protein is supposed to fold, this chaperone can, given enough time and energy, bias the folding pathway towards folding. It also seems to be able to detect that it contains a folded protein, prompting opening of the chamber.

Like stability, solubility is another danger. Proteins that are not sufficiently hydrophilic on their outsides tend to glomb together and create scrambled clots in cells, like we find in brains with dementia. Keeping proteins from aggregating is another job for chaperone proteins, which can pull such proteins apart, and / or tag them for degradation altogether.

A recent paper describes how chaperones can help to buffer organisms from the effects of modestly deleterious mutations, allowing a more diverse landscape of mutations in a population than otherwise possible. This is not a new idea, but the researchers try to put a more quantitative spin on it, and look in novel places for the effects chaperones have.
"For instance, enhancing chaperone capacity through over-expression has been directly shown to promote enzyme evolvability. Chaperones have been found to act both as genotypic and phenotypic capacitors."

Some of the most interesting bits of work concern protein interaction networks. Over the years, biologists have found that the cell full of proteins is a little like the internet, with lots of interactions between various proteins. There are some very dense sub-networks with a few hub proteins that have many partners, and then more peripheral proteins. These hub proteins tend to have a bit more disorder in their protein structure, perhaps because their many interactions require a bit of hydrophobic exposed area for each one (or else a floppy region that hides the interaction area prior to the partner's appearance). So these proteins are particularly dependent on chaperone assistance in folding:
"Upon deletion of [chaperone] SSB, 50% of the hub proteins, but less than 10% of the non-hubs aggregate immediately."
Hub proteins, by the author's analysis, tend to be more frequent "clients" of chaperones such as Hsp90 and Hsp70, and have more disordered regions in their structure.

They also do an interesting survey of the degree of rewiring of such protein networks between the distantly related yeast species, S. cerevisiae and S. pombe, to set up a metric of which proteins have maintained pretty much the same function, vs those whose partnerships have changed the most (i.e. become "rewired"). They then use two metrics of evolutionary change within protein sequences, the non-synonymous vs synonymous mutation rate, and second they devise a new conservative vs non-conservative change metric for mutations that are non-synonymous. This allows them to find that while highly rewired proteins tend to have lower than average non-synonymous mutation rates, befitting their key status in their networks, (i.e. increased purifying selection), they also have higher than average non-conservative amino acid mutations, perhaps as a sign of positive selection at key spots, plus, of course, the assistance that chaperones provide to enable structurally marginal mutants to survive.
"[Chaperone] Hsp90 was found to promote protein evolutionary rates in strong substrates [dependent on Hsp90 assistance for folding] when assessed by dN or ω [metrics of degree of change in protein sequences]."

Such analyses can help us dive ever deeper into the details of evolutionary history, given the vast resources of the genome sequences now available at every turn. In this case, it emphasizes that while each protein, indeed each nucleotide position, is an individual case, we can make some crude generalizations from sequence to function and back again. For the chaperones, it is evident that they are most important for difficult proteins, whether by intrinsic design in the interaction vs independent folding tradeoff, or temporarily after a damaging mutation, which may be resolved back to the original state eventually, or to some new state with new significance for the organism.


  • Confessions of a FOX youth.
  • Krugman on the prudential argument for higher interest rates: "No, what the BIS is arguing is that there is some other appropriate rate, defined as a rate sufficiently high to discourage bubbles, and that central banks should target this rate even though it is above the Wicksellian natural rate – or, equivalently, that the economy should be kept permanently depressed in order to curb the irrational exuberance of investors."
  • Much current international conflict revolves around energy.
  • Summers on our policeman role.
  • Ponzis are for pikers.. real frauds get into banking.
  • Hard money for old people.
  • The labor market should have been frictionless and efficient by now.. what happened?
  • Walgreens to move to Switzerland, sort of.
  • This week in the WSJ; nothing like the pot calling the kettle black: "Our political system is adept at making use of people like Mr. Steyer. Democrats will gladly spend his $100 million, then go back to their real environmental business, which is green cronyism."
  • Economic graph of the week ... how real & potential GDP has declined through our recession, for no reason other than gross macroeconomic mismanagement.

Saturday, July 5, 2014

Give a Guy a Hammer ...

Mathematician Max Tegmark thinks the fundamental reality is math. A review of Our Mathematical Universe.

The unknown seems to drive us into conniptions, whether one's habit of thought is theology, science, or formal philosophy. The idea that the fundamental reality of our cosmos might be inexlicable is as foreign to the most advanced scientist as it was to the earliest shaman.

So there we are. Physicists are knocking their heads against several walls such as dark energy, the proper interpretation of quantum mechanics, the union of quantum mechanics and relativity / gravity, and of course, the origin of the universe. They have virtually run out of experimental options, the colliders having become as super as they are realistically going to get. What now?

One can sense this fix from recent years of the magazine Scientific American, which runs ever more fanciful articles about the nature of the universe under the heading of physics. Speculation is running rampant, and the field seems to be gradually leaving the orbit of reason. What is time? What is space? Quantum foam, strings, etc.. All worthy questions, but far too speculative and sketchy to be fed to lay readers.

A recent entrant in the cosmic speculation derby is Max Tegmark with a book about how the universe is all a big mathematical structure. It is an excellent book in most respects, very readable and fair on the known science. Even sensible in a pontifical denouement of social policy. He has the most sterling credentials as an MIT physics professor, cosmologist, and protoge of John Wheeler. I should add that I am no expert in the least respect here, so I am just offering an educated lay perspective on the book and its ideas, as presented.

There are excellent aspects also to his cosmological speculations. For instance, he develops a helpful hierarchy of multiverse categories, this being a book largely about multiverses:

Level 1 multiverse: This is the notion that inflation during the big bang gave rise not only to the region of space we can see, but to much more. How much more? Hard to say, but it could be rather enormous, all within the product of the big bang we date to ~13.8 billion years ago.

Level 2 multiverse: Here the additional notion is added that inflation, the key process that we know of from the big bang, could have been a continuous process, not just producing our universe, but many, indeed an infinite number, of others in a process that is still going on. It adds the idea that these others might have different basic physics- different constants, symmetries, etc. Why this would be is due to the unboundedness of our current theories of what might have gone on. So why not everything possible?

Level 3 multiverse: Hugh Everett came up with an interpretation of quantum mechanics that contradicts the Copenhagen interpretation, and posits that the Schrödinger equation never "collapses". It just spawns other realities where events we think occur randomly actually occur in all possibilities, each in its own reality. This does not imply the multiplication of mass and energy into these other universes, but the superposition of an infinity of different possibilities in the mathematical space of quantum mechanics- the Hilbert space- of which we see only one sample at any moment. So it all looks the same as the Copenhagen collapse interpretation.

Level 4 multiverse: This is Tegmark's special theory, where not only does the level 2 multiverse generate an infinity of universes with different laws from some originating ur-structure, but even the most basic mathematical structure- his ultimate reality- can differ to generate alternate inflation (or non-inflation) regimes, of evey possible type. Indeed, he speculates that every computable mathematical structure exists and generates its own multiverse..es.

To be brief, I can easily understand the level 1 multiverse, and don't have a big problem with the level 3 multiverse of quantum mechanics. The others are a different story. Level 2 seems a cop-out, interpreting a lack of knowledge and specification about the universe as a permissive free-for-all where everything possible occurs. The premise is, as Tegmark notes, that our universe has about 32 numbers from which physicists can, in principle, calculate all physical aspects of our universe (not counting the pending conundrums of dark energy and dark matter, among others). And the values of these numbers are, of course, quite important. Any little change here or there would blow us to smithereens. So how did they get set up?

There are two basic approaches. The traditional way was to say god did it, end of story. A slightly more updated version is to look into the matter scientifically and keep hunting for simplifying and unifying theories, especially using mathematics. This has been the job of physics for several centuries, and seems to have arrived at a sizeable set of irreducible particles and forces, but can't seem to break through to a universal theory. The most modern way is to say that all the possibilities occur in all possible universes, of which there are an infinity, and we find ourselves, naturally, only in the one that lets glorious us happen. Ergo, the level 2 multiverse.

What is the prospect of yet more simplifying and unifying insights into the universe(es)? I have no idea. But the multiverse hypotheses seem to give up prematurely, and to what end? Even with a virtual infinity of universes, the chance that we get one that has 32 numbers, some possibly irrational, and thus almost impossible to get precisely right, ranging over countless orders of magnitude, still seems slim. So I am still rooting for a unifying explanation rather than a ramifying one whose sense is saved only by the anthropic principle. And that is really what we are talking about at this point- a rooting interest in where scientific speculation heads, since no evidence to date decides among these possibilities, and evidence may never do so.

Now we get to the weirdest part of the book- the level 4 multiverse, or Tegmark's theory that reality, at its base, is math, not just that it is described by math. And that all possible mathematical structures give rise to their own multi-multiversi, etc., ad infinitum. This is all more than a little fanciful. And his arguments, forming the core of the book and the armature around which so much else is built, are surprisingly weak.

The beginning premise is that external reality exists, separate from us, and even separate from us as observers. This is not at all hard to accept. After all, the universe had to roil and moil for quite some time before we were here to observe it, so the people who posit reality as a figment of our imaginations, or quantum-wise demand observation as the requirement of reality, do not have much to stand on. Then Tegmark goes on with the rest of his argument, which I abridge:

"If we assume that reality exists independently of humans, then for a description to be complete, it must also we well-defined according to nonhuman entities- aliens or supercomputers, say- that lack any understanding of human concepts."
... 
"This means that it [a master theory of everything] must contain no concepts at all! In other words, it must be a purely mathematical theory, with no explanations or 'postulates' as in quantum textbooks ..." 
"Taken together, this implies the Mathematical Universe Hypothesis, i.e. that the external physical reality described by the ToE [theory of everything] is a mathematical structure." 
"This means that our physical world no only is described by mathematics, but that it is mathematical (a mathematical structure), making us self-aware parts of a giant mathematical object. A mathematical structure is an abstract set of entities with relations between them. The entities have no "baggage": they have no properties whatsoever except those relations."

There, in a nutshell, is his argument. Note the slight of hand of getting from a description of reality to the reality itself. He explains himself later on:
"I'm writing is rather than corresponds to here, because if two structures are equivalent, then there's no meaningful sense in which they are not one and the same, as emphasized by Israeli philosopher Marius Cohen."

I can't say that this is convincing, at least to one untutored in the arts. One can also ask whether the starting premise makes any sense. Why must a universe be describable by any entities at all, human or non-human? It could just exist in some way and for some reason we can not understand or describe. The assumption is that there is a theory of everything, which I would certainly like to see. But I don't think it is a given that such a thing exists, let alone that it needs to have the describability property Tegmark claims for it. It could just as well be undescribable, and filled with the relatively arbitrary properties we actually see.

The one thing such a theory must be is consistent enough internally to produce a reality that has the symmetries and durable properties ("laws", constants, etc.) that we see in our versions of physics. And that, of course, is why mathematics is such a useful tool in physics, not because rocks are equations, but because our reality has, necessarily, the kinds of strucures and consistencies that we can use mathematics to describe. The ultimate theory may end up being a beautifully simple equation one can write on a T-shirt (as Tegmark dreams), but we don't know that yet, and it is very hard to see how that could be, with so many simple mathematical structures already known and tested in this respect. Are strings simple? Probably not.

And why one would want to theorize our reality as being a math structure ... that is admittedly beyond me. Tegmark claims that, among other benefits, this gets rid of an infinite regress issue, as we look for ever more fundamental particles and principles. (Though we have reached an end in particle terms, not being able to divide the electrons and quarks any further.) Having the most fundamental "one" be a total abstraction, and indeed every possible total abstraction in his level 4 multiverse, buys finality at the cost of nonsensicality, little better than the turtles or deities of yore. Specifically, it is Platonism revived, thinking that what is in our minds (where math is, exclusively) is the fabric of the universe, not its map. Indeed, one suspects in the end that this book is another edition in the old-as-humanity tradition of seeing the origins of the cosmos in the mirror.


  • The supremes are losing their minds. Hobby Lobby will live in infamy.
  • Can Muslim companies mess with their employees' healthcare and personal lives too?
  • The tortured reasoning of turning money into "free" speech.
  • Voters vote for climate action. How does money vote?
  • Bob Cringely: Bitcoins have come in from the cold.
  • Sectarianism, insurrection and theocracy ... not just somewhere else.
  • What money does to our minds.
  • On being a disposable worker at Walmart.
  • Bill Mitchell on European economic policy: groupthink followed by fiasco.
  • Jobs and the US economy.. have the green shoots finally arrived? We could have been here far, far sooner.
  • This week in the WSJ: "The more you help unemployed people, the more unemployed people you'll have."
  • This week in Das Capital: "Accumulation of wealth at one pole is, therefore, at the same time accumulation of misery, agony of toil, slavery, ignorance, brutality, moral degradation, at the opposite pole."
  • Economic quote of the week, from Joe Stiglitz:
"In fact, Geithner’s attempts to justify what the administration did only reinforce my belief that the system is rigged. If those who are in charge of making the critical decisions are so “cognitively captured” by the 1 percent, by the bankers, that they see that the only alternative is to give those who caused the crisis hundreds of billions of dollars while leaving workers and homeowners in the lurch, the system is unfair."