Thread by @BenCKinney

Alright, #SoCIA18 friends and followers. Next up is the conference's final act & second keynote: "Is there a sensible way to say Life is alive?" by Ford Doolittle of Dalhousie University!

Steps in the logic: 1. Life (capital L) = LUCA (Last Universal Common Ancestor) and her descendents. 2. Life and Life only (not lowercase-l) exists - see my tweets on yesterday's Carlos Mariscal talk...

...though the short version is that Life is better viewed as a cluster/individual than as an abstract natural category (which would be lowercase-l life). Anyways: 3. A Godfrey-Smithian approach to 'life' (whatever that means), 4.5.6.too fast.

Into the talk itself. Darwin thought life was originally "breathed into a few forms or into one" - i.e. the existence of a LUCA.

But, because of lateral gene transfer, LUCA could be an ancestrRY rather than an individual ancestOR. Basically, the tree of life's base trunk is broad, not narrow.

Okay hold on this image “explains” (or at least describes) the LUCA situation better than my words ever could.

It's possible that all genes of LUCA's descendants have been exchanged since then. But still, Life (the clade) is only LUCA + descendants: a very old and loose individual.

Ontologically speaking (metaphysical-nature-of-being), life on Earth is one individual. It exhibits continuity of processes all the way back to LUCA. This is a ~2016 scientific claim.

Whereas "species are individuals, organisms are paradigms" is mid-1970s. This is the Life/life thing I discussed yesterday/above: species are defined by continuity not characteristics. And have proper names.

"Eliminative pluralism:" conflicting species-concepts don't tell us that the taxa aren't real. The taxa are real, they just don't fall into one category ("species"). One individual taxa (that we call species, e.g. E.coli or humans), that identification is real.

Not sure I understood that last tweet, so it's ok if you don't.

All these definitions of life bump into the "N=1 paradox/problem" - we can't form a general universal theory of life, because we have nothing to compare Earth-life against. (Or so some people think!)

But our speaker (and a few others) think the "N=1 problem" is not a problem. Though the paper thereof is now undergoing a slew of academic inside-baseball peer-review problems

They combine individualism with eliminative pluralism, and say that Life (LUCA+descendnts) is an individual. Class/natural-kind "life" doesn't exist. If you still use "life/alive" (or "species") you must recognize it's arbitrary.

If there's no such thing as general "life" you don't need to worry about creating a general definition, so N=1 is no problem.

They think "life" might be a "parasitic kind"? It contains one individual who has all properties (ABCD) but also subsets that don't share anything with each other (A, BC, D). It's parasitic because once you lose the big ABCD there's nothing left the elements share.

Okay I *did* understand that one, can only hope I tweeted it clearly.

Another framing of their answer: Once we know the processes of a living organism (e.g. evolution, development, reproduction, metabolism), there's nothing left for the term "life" to add.

Forms of non-life: viral videos, Frankenstein's Monster, robots, life, the very last bunny - they all miss some of the components/criteria (e.g. don't participate in Darwinian selection, don't self-organize, etc)

Now we move onto "A Godfrey-Smith approach to life." He thinks there are two categories with special status: "organisms" and "Darwinian individuals." These groups overlap but are definitely not identical!

Organisms maintain structure despite turnover of material, by using energy + other resources from environment. Organisms are persisters, which resist entropy!

Darwinian Individuals change due to variation, heredity, and differing reproductive success. Can apply to genes, cells, social groups, and species. They must have capacity to reproduce? (Our speaker unsure about that last criterion.)

Many thinks are both Darwinian and Organisms. Us, flies, some symbiotic units but not all (aphid-Buchnera symbioses function as a single reproductive unit, but not squid-Vibrios)

Robots, last-of-kinds, and Frankenstein's Monster are organisms but not Darwinian. The reverse includes viruses, RNA, and viral videos.

Our definition of "life" (lowercase L here, for when we are forced to use this perhaps-useless term) is the overlap. Or, rather, things that are strong on both axes. (Each criterion can be a gradient, not just Y/N.)

Can we make it a 3d cube? Two axes are "organismality" and "Darwinian individuality," third axis is... we shall learn soon!

But first, an argument for organismality of Life! (Note the return to capital-L: the individual that is LUCA's descendents.)

The biosphere is analogous to a smaller organism's physiology. It has diverse parts that work together to maintain structure despite material turnover, using stuff from environment.

The harder question: is Life a Darwinian Individual? Life-as-a-whole doesn't seem to reproduce.

Similarly, the original counter-arguments against Gaia Hypothesis were that it doesn't reproduce/evolve. (A clade like Life by definition cannot have a parent-offspring lineage, because its descendants are part of itself!)

(Is Mars an example of a dead planet whose homeostatic regulation system failed? Will Earth be, after we're done with it?)

Our proposal here (and other people) are willing to ditch the Reproduction Requirement for a Darwinian individual. There can be selection advantages even without reproduction - but "fitness" leads to longevity and expansion, not to reproduction.

This idea (Clade Selection) is unpopular with philosophers. But he suggests that clade growth (proliferation of species in clade) substitutes for clade reproduction. (Reminder: a clade is a family tree and its ancestor, like Life.)

Clade-level properties favoring persistence – richness, dispersal, divergence, and possible intraclade cooperation – are not collapsible into species-level traits? This *expands* the explanatory power of Darwinian evolution.

Clade selection is based on emergent properties: yes it arises from lower-level things, but that doesn't mean it is the lower-level thing.

A clade can die/shrink because its species are too few, too geographically close, or too ecologically similar - these are clade-level deficiencies.

Paleontologists are constantly comparing one clade against another to learn why one succeeded over another!

Thus, we DID once have a population - the current situation is the sole survivor! At the time of LUCA there were probably competitors (other clades, dead branches of tree of life).

Evolvability is an evolved ability? The coding concept (i.e., ours is DNA-and-its-rules) enabled this biosphere to survive so long, including by altering the world environment, through at least 5 mass-extinction catastrophes.

Imagine a distant future where there are fewer surviving species - which means LUCA is much higher up on the tree (e.g. all species are descended from post-nuclear cockroaches). Other clades, now alive, will then be extinct. The clade evolves, continues, but different membership.

So this gives us the 3rd dimension of our criteria: reproduction -> persistence?

Our Criteria Cube might look like this...

Or maybe a little clearer on that third axis, like this. So maybe kinda we do have a way to say Life is alive. We knows where it fits on our crazy criteria cube!

Not to be confused with the simultaneous 4-day time cube.

Q&A time. Reproduction-Persistence is an axis because normal reproductive selection doesn't select for long life - any such is a side effect of the selection for successful reproduction.

It is possible to have a second individual who qualifies as life (a second "Life Individual"). Other worlds, for the obvious example. Our speaker is willing to call the Catholic Church another Life Individual. (Persistent, organismal, Darwinian.)

I'm not sure whether it's "Life individual" or "life individual" in this case (it's spoken, not on slide), but I think the former.

Note that biologicality is not one of the criteria. Self-replicating robots could be alive. He thinks the Catholic Church is probably alive, as noted. If you want to make biologicality another criteria, you need to set some (more) definitions. Which is hard.

How does this affect ethics? Moral consideration might track organismality well: we care about Frankenstein Monsters (+organism –reproduction) but not viral videos (–organism +reproduction)...

...In which case, how does this affect planetary protection? Possibly, the "persistence" axis might be another one that matters morally: like organismality, unlike Darwinianism. (More persistent things have more moral value?) We're all still pondering that.

How much does consciousness matter? This framework is very anti-vitalist (or doesn't care about vitalist questions), which can be a bit disconcerting.

The Q&A gave me a chance to photograph the “what if everything ends up descended from roaches?” tree of life.

Note that a clade (like Life) cannot "split," it can only widen/narrow/die. If you split, both branches have a common ancestor. Sending EarthLife to two planets greatly increases clade survivability.

Reproducing robots are certainly life (organized, Darwinian), but are they Life? Depends on whether we include "built by" in our relationships. Same goes for synthetic biology. A different but indeed interesting philosophical question.

Final comment: Darwinian populations produce most of the organisms we're interested in. Thus reason to morally care about the populations (Darwinian individuals)? There's certainly causal relationships between Darwinianism and organisms, but it's not logically certain.

Our conference organizer has just been described as an evil organism that looks and acts like a cat. He conceded the point.

And that's it for the official #SoCIA18 program! I'll make sure to collect and link every one of my bazillion threads, sometime tomorrow or maybe later in the week.

Thanks to all the awesome conference speakers & people, and everyone who's been following my threads! If you have questions for me, now I might have the time to actually answer them :)

unroll