Adam Victor Brandizzi

Why does life exist?

Popular hypotheses credit a primordial soup, a bolt of lightning, and a colossal stroke of luck.

But if a provocative new theory is correct, luck may have little to do with it. Instead, according to the physicist proposing the idea, the origin and subsequent evolution of life follow from the fundamental laws of nature and “should be as unsurprising as rocks rolling downhill.”

From the standpoint of physics, there is one essential difference between living things and inanimate clumps of carbon atoms: The former tend to be much better at capturing energy from their environment and dissipating that energy as heat. 

Jeremy England, a 31-year-old assistant professor at the Massachusetts Institute of Technology, has derived a mathematical formula that he believes explains this capacity. The formula, based on established physics, indicates that when a group of atoms is driven by an external source of energy (like the sun or chemical fuel) and surrounded by a heat bath (like the ocean or atmosphere), it will often gradually restructure itself in order to dissipate increasingly more energy. This could mean that under certain conditions, matter inexorably acquires the key physical attribute associated with life.

Kristian PetersCells from the moss Plagiomnium affine with visible chloroplasts, organelles that conduct photosynthesis by capturing sunlight. “You start with a random clump of atoms, and if you shine light on it for long enough, it should not be so surprising that you get a plant,” England said.

England’s theory is meant to underlie, rather than replace, Darwin’s theory of evolution by natural selection, which provides a powerful description of life at the level of genes and populations. “I am certainly not saying that Darwinian ideas are wrong,” he explained. “On the contrary, I am just saying that from the perspective of the physics, you might call Darwinian evolution a special case of a more general phenomenon.”

His idea, detailed in a paper and further elaborated in a talk he is delivering at universities around the world, has sparked controversy among his colleagues, who see it as either tenuous or a potential breakthrough, or both.

England has taken “a very brave and very important step,” said Alexander Grosberg, a professor of physics at New York University who has followed England’s work since its early stages. The “big hope” is that he has identified the underlying physical principle driving the origin and evolution of life, Grosberg said.

“Jeremy is just about the brightest young scientist I ever came across,” said Attila Szabo, a biophysicist in the Laboratory of Chemical Physics at the National Institutes of Health who corresponded with England about his theory after meeting him at a conference. “I was struck by the originality of the ideas.”

Others, such as Eugene Shakhnovich, a professor of chemistry, chemical biology and biophysics at Harvard University, are not convinced. “Jeremy’s ideas are interesting and potentially promising, but at this point are extremely speculative, especially as applied to life phenomena,” Shakhnovich said.

England’s theoretical results are generally considered valid. It is his interpretation — that his formula represents the driving force behind a class of phenomena in nature that includes life — that remains unproven. But already, there are ideas about how to test that interpretation in the lab.

“He’s trying something radically different,” said Mara Prentiss, a professor of physics at Harvard who is contemplating such an experiment after learning about England’s work. “As an organizing lens, I think he has a fabulous idea. Right or wrong, it’s going to be very much worth the investigation.”

Courtesy of Jeremy EnglandA computer simulation by Jeremy England and colleagues shows a system of particles confined inside a viscous fluid in which the turquoise particles are driven by an oscillating force. Over time (from top to bottom), the force triggers the formation of more bonds among the particles. At the heart of England’s idea is the second law of thermodynamics, also known as the law of increasing entropy or the “arrow of time.” Hot things cool down, gas diffuses through air, eggs scramble but never spontaneously unscramble; in short, energy tends to disperse or spread out as time progresses. Entropy is a measure of this tendency, quantifying how dispersed the energy is among the particles in a system, and how diffuse those particles are throughout space. It increases as a simple matter of probability: There are more ways for energy to be spread out than for it to be concentrated.

Thus, as particles in a system move around and interact, they will, through sheer chance, tend to adopt configurations in which the energy is spread out. Eventually, the system arrives at a state of maximum entropy called “thermodynamic equilibrium,” in which energy is uniformly distributed. A cup of coffee and the room it sits in become the same temperature, for example.

As long as the cup and the room are left alone, this process is irreversible. The coffee never spontaneously heats up again because the odds are overwhelmingly stacked against so much of the room’s energy randomly concentrating in its atoms.

Although entropy must increase over time in an isolated or “closed” system, an “open” system can keep its entropy low — that is, divide energy unevenly among its atoms — by greatly increasing the entropy of its surroundings. In his influential 1944 monograph “What Is Life?” the eminent quantum physicist Erwin Schrödinger argued that this is what living things must do. A plant, for example, absorbs extremely energetic sunlight, uses it to build sugars, and ejects infrared light, a much less concentrated form of energy. The overall entropy of the universe increases during photosynthesis as the sunlight dissipates, even as the plant prevents itself from decaying by maintaining an orderly internal structure.

Life does not violate the second law of thermodynamics, but until recently, physicists were unable to use thermodynamics to explain why it should arise in the first place. In Schrödinger’s day, they could solve the equations of thermodynamics only for closed systems in equilibrium. In the 1960s, the Belgian physicist Ilya Prigogine made progress on predicting the behavior of open systems weakly driven by external energy sources (for which he won the 1977 Nobel Prize in chemistry). But the behavior of systems that are far from equilibrium, which are connected to the outside environment and strongly driven by external sources of energy, could not be predicted.

This situation changed in the late 1990s, due primarily to the work of Chris Jarzynski, now at the University of Maryland, and Gavin Crooks, now at Lawrence Berkeley National Laboratory. Jarzynski and Crooks showed that the entropy produced by a thermodynamic process, such as the cooling of a cup of coffee, corresponds to a simple ratio: the probability that the atoms will undergo that process divided by their probability of undergoing the reverse process (that is, spontaneously interacting in such a way that the coffee warms up). As entropy production increases, so does this ratio: A system’s behavior becomes more and more “irreversible.” The simple yet rigorous formula could in principle be applied to any thermodynamic process, no matter how fast or far from equilibrium. “Our understanding of far-from-equilibrium statistical mechanics greatly improved,” Grosberg said. England, who is trained in both biochemistry and physics, started his own lab at MIT two years ago and decided to apply the new knowledge of statistical physics to biology.

Using Jarzynski and Crooks’ formulation, he derived a generalization of the second law of thermodynamics that holds for systems of particles with certain characteristics: The systems are strongly driven by an external energy source such as an electromagnetic wave, and they can dump heat into a surrounding bath. This class of systems includes all living things. England then determined how such systems tend to evolve over time as they increase their irreversibility. “We can show very simply from the formula that the more likely evolutionary outcomes are going to be the ones that absorbed and dissipated more energy from the environment’s external drives on the way to getting there,” he said. The finding makes intuitive sense: Particles tend to dissipate more energy when they resonate with a driving force, or move in the direction it is pushing them, and they are more likely to move in that direction than any other at any given moment.

“This means clumps of atoms surrounded by a bath at some temperature, like the atmosphere or the ocean, should tend over time to arrange themselves to resonate better and better with the sources of mechanical, electromagnetic or chemical work in their environments,” England explained.

Courtesy of Michael Brenner/Proceedings of the National Academy of SciencesSelf-Replicating Sphere Clusters: According to new research at Harvard, coating the surfaces of microspheres can cause them to spontaneously assemble into a chosen structure, such as a polytetrahedron (red), which then triggers nearby spheres into forming an identical structure. Self-replication (or reproduction, in biological terms), the process that drives the evolution of life on Earth, is one such mechanism by which a system might dissipate an increasing amount of energy over time.

As England put it, “A great way of dissipating more is to make more copies of yourself.”

In a September paper in the Journal of Chemical Physics, he reported the theoretical minimum amount of dissipation that can occur during the self-replication of RNA molecules and bacterial cells, and showed that it is very close to the actual amounts these systems dissipate when replicating.

He also showed that RNA, the nucleic acid that many scientists believe served as the precursor to DNA-based life, is a particularly cheap building material. Once RNA arose, he argues, its “Darwinian takeover” was perhaps not surprising.

The chemistry of the primordial soup, random mutations, geography, catastrophic events and countless other factors have contributed to the fine details of Earth’s diverse flora and fauna. But according to England’s theory, the underlying principle driving the whole process is dissipation-driven adaptation of matter.

This principle would apply to inanimate matter as well. “It is very tempting to speculate about what phenomena in nature we can now fit under this big tent of dissipation-driven adaptive organization,” England said. “Many examples could just be right under our nose, but because we haven’t been looking for them we haven’t noticed them.”

Scientists have already observed self-replication in nonliving systems. According to new research led by Philip Marcus of the University of California, Berkeley, and reported in Physical Review Letters in August, vortices in turbulent fluids spontaneously replicate themselves by drawing energy from shear in the surrounding fluid. And in a paper in Proceedings of the National Academy of Sciences, Michael Brenner, a professor of applied mathematics and physics at Harvard, and his collaborators present theoretical models and simulations of microstructures that self-replicate. These clusters of specially coated microspheres dissipate energy by roping nearby spheres into forming identical clusters. “This connects very much to what Jeremy is saying,” Brenner said.

Besides self-replication, greater structural organization is another means by which strongly driven systems ramp up their ability to dissipate energy. A plant, for example, is much better at capturing and routing solar energy through itself than an unstructured heap of carbon atoms. Thus, England argues that under certain conditions, matter will spontaneously self-organize. This tendency could account for the internal order of living things and of many inanimate structures as well. “Snowflakes, sand dunes and turbulent vortices all have in common that they are strikingly patterned structures that emerge in many-particle systems driven by some dissipative process,” he said. Condensation, wind and viscous drag are the relevant processes in these particular cases.

“He is making me think that the distinction between living and nonliving matter is not sharp,” said Carl Franck, a biological physicist at Cornell University, in an email. “I’m particularly impressed by this notion when one considers systems as small as chemical circuits involving a few biomolecules.”

Wilson BentleyIf a new theory is correct, the same physics it identifies as responsible for the origin of living things could explain the formation of many other patterned structures in nature. Snowflakes, sand dunes and self-replicating vortices in the protoplanetary disk may all be examples of dissipation-driven adaptation. England’s bold idea will likely face close scrutiny in the coming years.

He is currently running computer simulations to test his theory that systems of particles adapt their structures to become better at dissipating energy. The next step will be to run experiments on living systems.

Prentiss, who runs an experimental biophysics lab at Harvard, says England’s theory could be tested by comparing cells with different mutations and looking for a correlation between the amount of energy the cells dissipate and their replication rates.

“One has to be careful because any mutation might do many things,” she said. “But if one kept doing many of these experiments on different systems and if [dissipation and replication success] are indeed correlated, that would suggest this is the correct organizing principle.”

Brenner said he hopes to connect England’s theory to his own microsphere constructions and determine whether the theory correctly predicts which self-replication and self-assembly processes can occur — “a fundamental question in science,” he said.

Having an overarching principle of life and evolution would give researchers a broader perspective on the emergence of structure and function in living things, many of the researchers said. “Natural selection doesn’t explain certain characteristics,” said Ard Louis, a biophysicist at Oxford University, in an email. These characteristics include a heritable change to gene expression called methylation, increases in complexity in the absence of natural selection, and certain molecular changes Louis has recently studied.

If England’s approach stands up to more testing, it could further liberate biologists from seeking a Darwinian explanation for every adaptation and allow them to think more generally in terms of dissipation-driven organization. They might find, for example, that “the reason that an organism shows characteristic X rather than Y may not be because X is more fit than Y, but because physical constraints make it easier for X to evolve than for Y to evolve,” Louis said.

“People often get stuck in thinking about individual problems,” Prentiss said.  Whether or not England’s ideas turn out to be exactly right, she said, “thinking more broadly is where many scientific breakthroughs are made.”

Emily Singer contributed reporting.

This article originally appeared at Quanta Magazine. Copyright 2014. Follow Quanta Magazine on Twitter.

I have $10 that says you’ve experienced this before: there’s a holiday, trade show, or other important event coming up. Management is worried about the risk of an outage during this all-important time, and restricts deployments from the week prior through the end of the event.

What really happens, of course, is that the system in question becomes booby-trapped with extra risk. As a result, problems are more likely, and when there there is even a slight issue, it has the potential to escalate into a major crisis.

Why does this happen? As usual, there’s no single root cause, but a variety of problems combine to create a brittle, risky situation.

Assumptions

When managers declare a freeze, they’re not being malicious. They’re doing something that seems to make sense. That’s why it’s important to understand the reasoning.

The goal is simple: prevent breaking something that’s working. And a lot of issues seem to trace back to a deployment that breaks a working system, so it’s natural to want to avoid a deployment.

The assumptions, in my experience, are that

• the systems are working now
• systems break because of change
• restricting change will prevent breakage
• the riskiest change is deployment, so restricting deployment eliminates risky change
• potential downsides are less costly than potential outages

Unfortunately, the assumptions are wrong, which is why deployment freezes often have the opposite of the intended effect.

Change Breaks Things

It’s true that change breaks things, but that doesn’t tell the whole story. The reality is that systems are always changing, even when you think they’re not. And they’re also broken. Right now. You just don’t know it yet.

The first great lie is that you can stop systems from changing. You can’t stop change. Shipping code is not the only change in a system. Unless the system is turned off completely, it’s changing every instant. User requests are being served. Logs are being written. Cron jobs and other scheduled tasks are starting, running, and stopping. The system clock is advancing. There are also dependencies on other systems, which likewise are undergoing constant change. And there are requirements changes, which nobody ever thinks about.

Most of these changes happen all the time and nothing goes wrong, or humans adjust and fix things as a matter of course. But many of these changes actually expose corner-case bugs or weird problems that just aren’t tickled often. As a result, you get the illusion that your systems are not broken. And that’s the second great lie, because your systems are broken. You just don’t know it yet.

What’s broken? All sorts of things. Your Java runtime has a time bug that’s waiting to throw it into an infinite loop when a leap second happens. Your website has a SQL injection vulnerability that nobody’s exploited yet. A set of servers that you bought all together and installed at the same time has a ticking time bomb in the RAID controllers, which are all going to go into degraded self-test mode simultaneously. Your auto-increment integer column is crossing the 2 billion mark right now, and will soon reach its max value.

The reality is that practically every running system is a ticking time bomb. You just don’t know what’s ticking or how long till it stops.

Uncontrolled Change Is More Dangerous

As I said, I’m not demonizing your managers. They’re actually right to think that deployments are riskier and are likely to cause breaking changes. That’s because deployment is often a barely-controlled change.

Generally, the more automated (scripted) the deployment is, the safer. This is actually not because of the automation, but because of an effect of automation. With automated deployment in place, deployment usually becomes a more frequent process, which tends to exercise, expose, and remove sources of risk to a larger extent.

So the headline of this section is wrong, in a way. Uncontrolled change isn’t what’s dangerous. Infrequent change is the problem.

Wait! Isn’t that precisely the opposite of what the managers are asking? Aren’t they asking us to freeze deployments in order to slow the pace of change and decrease risk? Yep, that’s right. Their intuition is actually the reverse of the truth: to make the systems safer, they should be encouraging more deployments, not fewer.

Deplyoment freezes actually add risk. But that’s not all. The deployoment freeze sets in place a vicious cycle that includes a couple of other effects and spirals down and down. Watch as I show you how the deplyoment freeze is only the first step in how we ambush the operations engineers.

You Can’t Freeze Deployments

Freezes never, but never, work. You can declare a freeze, but you can’t make it happen.

This follows naturally from the inability to actually stop the systems from changing. Systems that change are going to break, and you have to respond to this.

There are other effects, too. Requirements change, for example. The catalog merchants went to the marketers and decided to offer a special promo. Buy a tablet and a case in combination with a warranty, and get a bluetooth keyboard free! Sounds great; we’ll make a bundle – let’s print and ship those catalogs. Just in time for the holidays! Whoops, nobody told IT about this. Catalogs are printed already. We’ve got two weeks and there’s no going back; we need the order-entry system and the website to support this promo. Bingo; forced change.

Whether it’s a promo or a bugfix, something always has to be changed. So we always have to break the rule. This is why the rule inevitably ends up being no deployments except for emergency changes.

This is getting good. What’s riskier than a change? An emergency change. What’s riskier still? Undeployed code!

Undeployed Code Is Inventory; Inventory Is Risk

The process of developing and deploying code is really a process of forking and merging your codebase. A developer writes a new version of the application, which diverges from what’s running in production. The developer then merges it back in and deploys it to production, resolving the differences.

There are at least two key points where inventory of risk builds up in this process. The first is when the developer’s codebase isn’t merged into the main code yet. The longer this waits, and the more the code diverges, the riskier. Secondly, after merging and before deployment.

Code that has been merged and not deployed is a loaded gun. If I merge in my changes and don’t deploy them, and you then merge and deploy yours, you’ve just deployed mine too. This was more than you bargained for. It’s now more likely that your deployment will break something, and harder for you to fix if it does.

In a deployment freeze, two important things happen.

1. Lots of latent changes build up, ready to break the whole world when the freeze is lifted. There is increasing risk of breaking the system after the freeze.
2. Every developer’s codebase, and knowledge of the codebase, is diverging from production. Development and QA environments typically are, too. As a result, when emergency changes have to be applied to production during the freeze, breakage is much more likely. There is increasing risk of breaking the system during the freeze.

Frozen Systems are Inoperable

We’ve seen that the tip of the codebase is where the least risk has accumulated. The tip is where things work best. Frozen systems get further behind this tip, and therefore much riskier.

Infrastructure is code, too. Frozen systems also become incompatible with the current tip of environment configuration and infrastructure automation code. The result is that emergency deployments are often done outside of normal change control procedures, and the usual automation and deployment code can’t be used.

To illustrate this problem, consider that app code and automation code are usually separate, although there is a dependency. This dependency is not properly versioned most of the time. Is your Puppet code version-controlled in lockstep with your application code, such that if you want to deploy a stale version of the app code, the deployment process will check out and use the older Puppet code? Would this even work, if it were possible – can you just run old Puppet code on systems that have been advanced forward? The answer to both of these questions is likely no.

Imagine trying to drive your car without all the usual help you get from it: no mirrors, no gauges, no power steering, no power brakes. A car with all these things disabled is largely inoperable for most people.

A system that’s stale in production for a month, while developers work actively on a bunch of unshipped changes, is in a similar state. Important controls, indicators, and assists are disabled. It’s largely inoperable.

We’ve seen that a deployment freeze can’t prevent the need for changes. Now we see that it forces the inevitable changes to be done in a much more dangerous way, seriously increasing the risk of problems. Looks like the freeze is not preventing the risk of an outage as intended!

The Cost Of Deployment Freezes

The assumption is that the downsides of a deployment freeze are outweighed by the benefit of avoiding outages. I’m arguing that the benefit is much smaller than expected, and may even be negative. What about the costs?

Deployment freezes have a huge productivity cost. Huge.

When you can’t deploy, you can’t resolve the temporary tension in the system that results from changes that aren’t merged or deployed. Many changes you make while developing the system have to go through a compatibility lifecycle: to mutate from state A to B, you must first go to AB, which is backwards compatible with A and forward compatible with B. Then you move from AB to B.

A simple example is adding a column to a table. You make sure the code ignores new columns, add the column with a default value, then make the code recognize and work with the new column, then remove the default value once it’s no longer needed.

This kind of compatibility lifecycle represents a set of dependencies across time. If you’ve read The Goal, you know dependencies are bad for workflow. These dependencies become blockers for developers working on the code, and even for other developers and operations staff. If the dependencies are short-lived, it often isn’t too bad. This is the case when you deploy a dozen times a day. But if the dependencies become long-lived, large portions of engineering slows down or stops completely. Projects have to be shelved until things can be deployed. Meanwhile, work that’s been done on these projects becomes stale and creates conflicts with alternative work. It often has to be reworked later.

What was a local change and dependency becomes, given a little more time, a global blocker and waste of work.

This scenario is practically certain to happen. Deployment freezes crush productivity, often long beyond the duration of the freeze itself.

What About Feature Freezes?

It’s fairly common for large, complex, fast-moving systems to declare feature freezes in preparation for a major release. (Think of versions of operating systems or databases, for example.) However, this is different from a deployment freeze in a web appication.

The most important difference is that a deployment freeze is an attempt to branch the infrastructure, code/app, data, system state, and activity (customers/users). This is not branching one thing, but many interdependent things. This is practically impossible to do.

Conclusion

Frozen systems can run as-is briefly, but then pressure to change them mounts steeply. The system grows stale, but things change anyway and force changes. These changes become much more difficult and dangerous, and the likelihood of an outage grows quickly.

Code freezes thus make systems more likely to break, while impacting productivity. Both the cost and the risk are much worse than intuition would suggest.

I believe that a focus on improving the processes by which changes are made, making smaller and more frequent iterations, and finding out and fixing breakage as soon as possible is a better way to the goal. But that’s a different article

What have your experiences been? Let me know in the comments.

Image credits: freeze, clock, raccoon

A focus on “existential” or “categorical” inequality is what in the 19^th century Europe used to be called a radical position, associated with the post-1789 developments. Once all formal distinctions of class between clergy, aristocracy and people were abolished, there was, it was argued, no need to focus on the existing income differences. This view reached its peak under the French Third Republic when inequality was increasing by leaps and bounds, while formal equality was left untouched.(The socialist position at the time was that formal equality is just the first step towards real equality which requires also the diminution of economic inequalities.) The same radical position holds fast and true today: once you see the world as primarily composed of various groups, you quickly slip into “identity” politics whose main objective is to equalize formal legal positions of the groups—and basically let everything else the same.

According to Thorborn, that’s what the world has been remarkably successful in doing in the past 30 years. There are well-known and substantial advances in the equal treatment of different groups (listed above); there was also a strong push for “horizontal” equality, which is the term used in economics to indicate that on average there should be no wage differences between men and women, blacks and whites etc. (that is, atleast no differences that cannot be explained by better skills or experience). The progress there, although not as substantial as in legal equality, has been real too.

But the quasi single-minded focus on “existential” inequality was not always helpful, and I think in some cases was outright harmful, to the general reduction of income and wealth inequalities. The success in the latter would—I think it could be argued—also reduce income differences due to racial or gender discrimination. In other words, pushing for reduction of inequality in general would make lots of sense even if our primary objective is to reduce specific gender or racial income inequalities. But this is not how things worked out. Rather, the focus was on “horizontal” inequalities while the overall, general inequality was left to its own devices, namely was allowed to increase.

The focus on “existential” inequality is wrong, in my opinion, for at least three reasons. 

First, the emphasis of group differences quickly spills into identity politics, splintering the groups that do have an interest in fighting for change. The joint front crumbles. The groups end up by caring just about the change in their own positions and become indifferent to the rest. I am unaware for example that gays or immigrants, once their objectives of legal equality achieved, have shown particular interest to fight for economic equality in general, be it in the United State or the world. Splintering has made people focus on their own complaint; once that complaint is solved, they are indifferent to the rest. 

Second, the focus on “existential” inequality leaves the basic problem unsolved because the way it poses the issue is wrong. I noticed this in a recent discussion regarding legalization of prostitution. For feminists, prostitution is a reprehensible activity that they would like either to ban, discourage through some ill-defined teach-ins of women, or curb its demand by punishing clients who are predominantly males. Not only do these approaches just drive the problem underground without solving it, they are futile because the root case of prostitution is not addressed. The root cause today (and perhaps in history) is income and wealth inequality. There are many (mostly) men with huge incomes and there are many (mostly young) women with poor job prospects and no money. This drives prostitution nationally and globally (as in sex tourism). So,the point is not to address gender inequality only (men vs. women) but its economic cause. Consider what would happen even if horizontal equality between men and women were achieved, a thing which, with higher enrollment and graduation rates among women than men, and rising number of rich women, man soon happen. The problem will simply become that instead of 90% of customers being men we shall have a “fair” and “gender-neutral” distribution of customers with 50% men and 50% women. Will such gender equality solve the problem? Obviously not: prostitution, a reprehensibly activity in the eyes of the gender-focused activists, will merely become gender-balanced. Is this all they really want to achieve? No. But, of course, it reveals that the real cause of the problem lies elsewhere, in inequality, and that their approach is misdirected.

Third, the emphasis on “existential” equality is politically easy because it is not serious. It faces no real opposition from the right-wing politicians and conservatives because it does not affect the underlying structure of economic inequality and political power. Instead of fighting for meaningful general changes (e.g. increased vacation time for all, shorter work-week for all, more favorable working conditions for parents, longer maternity and paternal leave, higher minimum wage for all etc.)—the issues on which the success has been quasi nil, but which would cut into the profits and thus face a strong economic opposition from the businesses, proponents of the “existential” equality care only up to the point where legal equality is established. Strictly speaking, such equality is also in the well-understood interest of capitalists. We know at least since Gary Becker that, economically, discrimination is inefficient for those who practice it. But general measures that improve the position of the employees will not of course please those who have economic power. So the proponents of “existential” equality stop midway again. Formal equality is surely a necessary condition for overall betterment, but it is not sufficient. A movement toward more generalized equalization of human condition requires not only legal equality but also substantively greater (income and especially wealth) equality.

Their approach (“formal equality and then nothing”) is what Rawls calls “meritocratic equality”, the lowest level of equality where all participants are legally free to pursue whatever career they choose but where their starting positions are vastly different. All those who care exclusively about “identities” do only that: they aim to place everybody on the same starting line, but do not care if some come to the starting line with Ferraris and others with bicycles. Their job is done once everybody is on the same starting line. Case closed: just when the real issues begin.