by Ursula Goodenough

Stu Kauffman offers us below a summary of contemporary work on the origins of life. In the comments that followed, Ted Pawlicki posted the following  question:

Before Watson and Crick, people hypothesized a kind of force ("elan"
"chi" "spirit" "life force") that provided impetus for life. Does the
understanding of life as a chemical process eliminate the need for
this? Since there is no qualitative difference between the
chemical processes in "life" and "non-life" (the only difference seems
to be one of complexity), then why keep a term around if it's not
adding anything?

I give a lecture on the origins of life in an evolution course for non-science majors. My thinking is deeply influenced, as usual, by Terry Deacon; he’s posted a 13.7 teaser on the topic here; a free download of a relevant paper can be found here; and a full presentation of his ideas is forthcoming in his book Mind from Matter: The Emergent Dynamics of Life. Much of his thinking has roots in Stu Kauffman’s pioneering work on emergence.

Here’s how I tell my students the story, where I make use of two concepts: constraint and emergence.

We start with atoms –- each with a distinctive positively-charged nucleus and a distinctive array of negatively-charged electrons (think periodic table). Atoms can just bounce off each other, but under certain conditions, certain atoms will interact with one another such that their electrons form bonds. Bonds sometimes form spontaneously, like H + H, but bond formation usually requires work/energy input. The resultant entities are called molecules, production of molecules is called chemistry and the conditions that govern whether or not chemical bonds will form are dictated by the laws of thermodynamics.

Molecules of interest to our story have two properties: 1) the positions of their electron orbitals are constrained by bond formation such that their overall charge distributions are asymmetric; and 2) the positions of their nuclei are constrained such that each molecule has a distinctive shape. Collectively, these features can be said to arise via thermodynamic constraints.

 

Molecules can just bounce off each other, but under certain circumstances they interact via plus-to-minus charge interactions and/or shape-shape interactions (e.g. a protuberance fits into a cavity). These relationships add additional constraints to the positions of the nuclei and hence generate yet more complex shapes; these shapes can go on to interact via charge and shape to generate yet more complex shapes. Collectively, these features can be said to arise via morphodynamic constraints.

A familiar example of a constrained system is ice. Its individual water molecules each adopt a dipolar V-shaped configuration via thermodynamic constraints, and they then interact via mophodynamic constraints to form a crystalline lattice.

And now we can introduce our second concept: emergence. The arrangement of water molecules in an ice crystal leaves small interstices, whereas the water molecules in liquid water are statistically cheek-to-jowl. Hence ice is less dense than water, so it floats. The constraints that generate the crystal also generate an emergent property, namely, buoyancy. Buoyancy is not a property of individual water molecules. It arises, or emerges, or “pops through” only when water molecules engage in morphodynamic relationships at cold temperatures.

Emergent properties arising via thermodynamic and morphodynamic constraints are abundant in non-life: the hardness of minerals; the malleability of metals; the surface tension of liquids. To make the move from non-life to proto-life, we need to add one more thing: emergent properties that have functions, functions that are subject to natural selection.

At this point in the story I usually project a bunch of slides showing the functions achieved by a hypothetical proto-life entity called an autocell. Figure 1 of the Deacon free-download click has a great autocell cartoon if you like visuals. But let’s see how far I can get with language.

The key feature of any life form is that it is a self, a concept I introduced in an earlier blog.  A self engages in self-maintenance, and self-maintenance is accomplished by functional activities –- e.g. the ability to acquire energy resources, to produce molecules that catalyze chemical reactions, and to build a surrounding capsule that keeps molecular products from diffusing away. These functions are then evaluated by natural selection: If they succeed in maintaining a self, then the self continues to exist; if they fail to maintain a self, then the self disintegrates.

Critically, functions are emergent. When individual molecular units are morphodynamically constrained via self-assembly into a capsule, the emergent property is called self-containment. When individual enzymatic units are morphodynamically constrained into a recursive cycle as Stu describes, the emergent property is called autocatalysis.  A capsule that briefly disassembles and then reassembles to allow substrates to enter the self displays the emergent property/function called acquisition of energy resources. A capsule that disassembles and then reassembles around two independent autocatalytic sets engages in the emergent function called self-reproduction.

Autocells can evolve by encountering and acquiring novel substrates that become constrained into novel shapes with novel functions, but the big step in going from autocellular protolife to life was the invention of replicable instructions for self-generation in a separate molecular medium like DNA. This move allowed novelty to arise by mutating the instructions rather than by waiting for novelty to show up in the substrate pool.  Encoding was a great idea, and once it was in place, biological evolution could really take off. But it’s an add-on. The core origin-of-life event, in the Deacon model, is getting something like an autocell to maintain itself in its environmental context.  The probability of this happening, to be sure, is small. But there was lots of time back then.

So we can now re-visit Ted’s questions. Life is different from non-life because it generates selves with teleodynamic constraints, molecular arrangements that are for something, have a purpose, point to goals that, if achieved, allow the self to make the crucial natural-selection cut. Snowflakes and autocells both arise as the consequence of thermodynamic and morphodynamic constraints, but snowflakes aren’t self-maintaining, have no telos, aren’t “for” anything. Their emergent patterns are the passive consequence of temperature, pressure, humidity, and rate of fall. The emergence of an “élan vital,” a “life force” was coincident with the emergence of function.  The underlying chemistry of non-life and life is the same, but when functions arise and are selected, that changes the name of the game.