3.
What is the problem of biological
individuality? Compare and contrast two theories of biological individuality,
and explain whether they are successful or not.
The “problem of biological
individuality” is the problem of how to give a general account of what defines
a ‘biological individual’ across all species to which evolutionary theory
applies, given the incredible diversity of life. Ellen Clarke [2010] holds that
this is a serious issue because of how crucial the notion of an organism is to
our understanding of biological evolution, the very concept of “fitness”, and
more. In this essay, I will compare Peter Godfrey Smith’s elaborate, continuous
account of biological individuality with Ellen Clarke’s own, much simpler functional
(and also continuous) account of biological individuality. I will suggest that
they are both very impressive taxonomical efforts, but that Godfrey-Smith’s has
the greater balance of virtues, on account of (what I believe to be) its
greater comprehensiveness and greater potential to defuse controversies.
In her well-cited 2010 paper “The
Problem of Biological Individuality”, Ellen Clarke concisely and effectively explains
the titular problem, and why one should care. She begins the paper by explicating
the centrality of this idea of ‘the organism’ to the biological sciences. This
notion of the biological individual is perhaps most important for its starring
role in our understanding of evolution: Darwin formulated his theory of
evolution in terms of biological individuals, the “received view of biological
evolution takes the organism as “the basic unit of selection””, and, as Dawkins
admits, even the ‘gene-centrist’ cannot hope to dispense with the biological
individual in evolutionary theorising (even in the mathematics) [Dawkins, 1982:
251]. As Clarke puts it: “It is hard to overemphasize the importance of
individuals within the Modern Synthesis. They are central to the inner logic of
evolution by natural selection, according to which evolution occurs because of
the differential survival and reproduction of individuals” [2010: 313]. The
organism also plays a massive role, albeit slightly more hidden, in various
other sub-fields of the biological sciences “such as medicine, developmental
biology, immunology, ecology, and the reductionist sciences such as molecular
or cell biology” [313]. Finally, organisms are what population biologists
count!
The trouble,
of course, is that, despite the immense scientific utility of this concept, philosophers
of biology as of 2010 only had a long list of competing criteria for describing
biological individuality, all of which individually seem to admit of
counterexamples and define strongly “non-overlapping classes” – and this
vagueness allows for scientific conflict as well. Clarke discusses, in
particular, 13 different candidates for criteria (all biological “properties”
in some very wide sense) which could do the work of “differentiating
individuals from parts and groups”. [1]
Although some are more promising than others, the case studies Clarke
introduces demonstrate (I think) that there is not a clear mix-and-match
solution. In her follow-up 2013 paper
“The Multiple Realizability of Biological Individuals”, Clarke also shows that the
problem has practical import by pointing to some of the scientific controversies
that could have been avoided if biologists had agreed on what counts as an
individual. The first controversy she cites is the “long-standing debate
amongst plant scientists about whether vegetatively produced plants […] ought
to add to the parent plant’s fitness or not” [414]. Another, more general one she
describes is the long-running controversy in evolutionary theory over ‘levels
of selection’: in particular, over propositions like “selection always acts at
the level of the individual”. She suggests – rightly, I think – that these
debates would dissipate a lot of heat if the interlocutors acceded to a common
account of biological individuality (especially to deal with what Godfrey-Smith
calls the “problem cases” of collective entities like “ant and bee colonies,
and lichens” [Godfrey-Smith, 2012: 3]).
In the same 2013 paper, however, Clarke goes on to
propose a solution to her own problem. Her big idea is to ‘compress’ several of
the competing properties she highlighted in “The Problem of Biological
Individuality” (sex, bottlenecks, germ-soma separation, policing mechanisms,
spatial boundaries, and immune response) into a simpler and more economical
‘functional’ definition. The first component of Ellen Clarke’s functional
definition is the “policing mechanism”, which she claims is a robust enough
functional property to constitute a necessary
condition for biological individuality. She defines a policing mechanism as
“any mechanism that inhibits the capacity of an object to undergo within-object
selection” [2013: 421]. Clarke thinks that as well as helping to end debates
over the priority of this or that specific policing mechanism, this kind of
functional definition can put us in a better position “to recognize real-life
structures that play the desired role” [422]. She justifies this claim by
giving a number examples of such real-life structures, including “Resource
exchange, synchronized/vertical transmission (especially “co-dispersal”),
spatial contiguity or engulfment, the immune system, maternal control of early
development, and worker policing” [423].
Of course, whilst having a “policing mechanism” is a
necessary condition for something to be a biological individual, Clarke
recognises that it is by no means sufficient, on account of its being a
“negative” condition: for example, the non-organism that is a human muscle cell
“has ample policing mechanisms to cement common purpose amongst its component
organelles and genetic material” [423]. The necessary positive mechanism for
Clarke then is the “positive capacity to undergo natural selection at its own
level” [423]. After relating this suggestion to the biological function of sex,
Clarke extends this insight into a second functional criterion for biological
individuality: a “demarcation mechanism”, which she defines as “any mechanism
that increases or maintains the capacity of an object to undergo between object
selection” [424]. Like policing mechanisms, demarcation mechanisms are highly
multiply realisable. For example, spatial boundaries and immunity can often
play a ‘demarcating’ role. Demarcation, Clarke claims, is also “essential to an
evolutionary transition” [426]. As she
explains, “Evolutionary transitions in individuality can be viewed as a failure
to meet the demarcation challenge on the part of the lower-level individual.
Mitochondria, for example, have lost their biological individuality because
they became subsumed within eukaryotic cells” [426].
Clarke’s two mechanisms are, she thinks, sufficient
criteria for biological individuality. She holds that “Biological individuals
are all and only those objects that possess both kinds of individuating
mechanism” [427]. This may seem like a very bold claim, but Clarke is anxious
to point out is that the very nature of these mechanisms ensures that individuality
is a continuous concept. She makes
the important observation that “by incrementally increasing an object’s
capacity biological individuals for heritable variance in fitness, compared to
the capacity of its parts, individuating mechanisms can gradually push the
object through an evolutionary transition in individuality” [430]. Indeed, she
argues convincingly that it is necessary to recognise this kind of continuity
in individuation in order to understand how evolutionary transitions happen at
all.
In his 2012 paper “Darwinian Individuals”, Peter
Godfrey-Smith outlines a somewhat more complicated account of biological
individuality than Clarke’s. Although it has many features in common with
Clarke’s, and whilst it’s not clear that the two accounts are in any kind of
strong tension, Godfrey-Smith’s separation of two partly-overlapping sub-genres
of biological individuals, “Darwinian individuals” and “Organisms”, and his
very detailed accounts of each, lead to a scheme which is extremely good at
dealing with ‘borderline’ cases, and yet produces much more definite
entailments about specific cases than Clarke’s account. Clarke, I think,
effectively argues for the value of a ‘functional’ definition of biological
individuality, but I will argue that Godfrey-Smith’s work shows that a broadly
functionalist approach can be combined with specific biological properties to
produce a more complete overall account of biological individuality.
Both Clarke and Godfrey-Smith are chiefly concerned
with coming up with an account of biological individuality fully in tune with
the usage of “individual” within evolutionary theory. Clarke’s account is
peculiarly devoted to what Godfrey-Smith specifically demarcates as the
“Darwinian individual”, since her two individuating mechanisms have the
ultimate function of enhancing “heritable variation in fitness”. In
Godfrey-Smith’s attempt to come up with an account of the Darwinian individual,
reproduction is the key factor
(whereas for Clarke it goes along for the ride to some extent). Like Clarke,
however, Godfrey-Smith is mainly concerned with “collective” individuals in
formulating his account. Unlike Clarke, Godfrey-Smith’s account is explicitly continuous: he sees Darwinian
individuality in terms of three dimensions, with the most exemplary cases
measuring ‘high up’ in all three and non-Darwinian individuals measuring very
‘low down’ in all three. The first dimension or “parameter” is B, which stands for “bottleneck”. By
this, Godfrey-Smith means any kind of “narrowing” that “marks the divide
between generations […] often to a single cell” (as in humans) [2012: 6]. The
second dimension is G, which stands
for “germline”. G measures the degree of reproductive specialization within a
collective. This property helps usefully distinguish kinds of ‘eusocial’ species.
For example, in honey bee colonies “the queen reproduces (along with the male
"drones"), and the female workers do not” (high score for G), whereas
“In other insects, including other bees, there is no reproductive division of
labor” (low score for G) [7]. The third and final, more functional dimension is I, which
stands for “integration”. This does the ‘work’ of parts of both of Clarke’s
individuating mechanisms, involving a “general division of labor (aside from
that in G), the mutual dependence of parts, and the maintenance of a boundary
between a collective and what is outside it” [7]. Godfrey-Smith uses this
three-dimensional account to come up with an ingenious 1×1×1 cubic visualisation of
where various species ‘sit’ in terms of their level of Darwinian individuality.
Humans (and other mammals, marsupials, birds, many amphibians and many fish)
are prototypical Darwinian individuals, with perfect (1,1,1) scores for each
parameter; the Volvox carteri alga also
scores highly with 1, 1, 0.5 (B, G, I); clonal colonies like the Aspen ramet
scores 0.5, 0.5, 1; sponges score 0, 0, 0.5; and a buffalo herd is not a
Darwinian individual at all, since it scores 0, 0, 0.
It seems to me that Godfrey-Smith’s ability to
represent his classifications so elegantly represents a distinct advantage of
his account over Clarke’s. In this, I fully endorse Daniel Dennett’s praise of the
same diagrams in his review of Godfrey-Smiths’s 2009 book Reflections on Darwinian Populations and Natural Selection.
Godfrey-Smith then moves onto his account of
‘organismality’, where an ‘organism’ is understood as something distinct from a
Darwinian individual – a concept that Clarke does not have. The way
Godfrey-Smith defines an organism is as follows: “Systems comprised of diverse
parts which work together to maintain the system's structure, despite turnover
of material, by making use of sources of energy and other resources from their
environment” [12]. This he calls the traditional, ‘metabolic’ view of a
biological individual. Ultimately, the combination of this continuous organism
concept and the Darwinian individual concept allows for Godfrey-Smith to define
biological individuality for all of life.
Many biological individuals – like humans, or fruit flies – are both
Darwinian individuals and organisms.
A much smaller number would be classified as relatively prototypical examples
of Darwinian individuals but not organisms: “scaffolded reproducers” like
viruses, along with chromosomes and genes [16]. Finally, some organisms are not Darwinian individuals.
The more significant cases in this category “are certain kinds of symbiotic
associations” [16]. Godfrey-Smith cites Dupre and Malley [2009] as showing that
“most or all plants and animals live in close associations with symbionts” [16].
One specific example of a very close symbiotic relationship whose significance
has only recently been uncovered is that between various tree species and “mycorrhizal
fungi” which connect root systems in forests such that trees can ‘communicate’
threats and distribute resources to other trees in stress [Macfarlane, 2016]. (Basically, such tree-fungi fusions seem to fall under the category of organisms that are not Darwinian individuals (the fungi are crucial to the tree's fitness, as in the other example of a symbiont that is an organism but not a Darwinian individual which I'm about to discuss, but the tree and the fungi do not reproduce together, as one, in contrast to the aphid-Buchnera symbiosis.) Godfrey-Smith’s best example of an organism (albeit a
non-prototypical organism) that is not a Darwinian individual is the
“squid-Vibrio combination”, which has a “horizontally transmitted symbiont” as
opposed to the “vertically transmitted symbiont” of the oft-cited aphid-Buchnera
symbiosis. Even though the squid has evolved
six internal ‘chambers’ designed to take in the bacteria that create a
luminescent, moon-light-like patterning on their body and help them avoid avian
predation at night, the fact that the squid are not born with these bacteria
inside them means the partnership does not count as a Darwinian individual.
I think the big advantage Godfrey-Smith’s complicated
account has over Clarke’s much more economical one is that, whereas he can
apply his scheme to these exotic cases and produce definite (albeit
‘continuous’) verdicts, such verdicts do not directly fall out of Clarke’s considerably
looser scheme (it seems to me that Clarke's account makes it very hard to disentangle the very relevant differences between the type of symbionts I discussed, for example). Clarke, in fact, ends her 2013 paper by insisting on the
implausibility of a general system of classification for all of life – and yet
it seems to me that that’s effectively what Godfrey-Smith achieves.
Reference
List
Clarke, Ellen
(2010). “The Problem of Biological Individuality”, Biological Theory, 5 (4): 312-325.
(2013). “The
Multiple Realizability of Biological Individuals”, Journal of Philosophy 110
(8): 413-435.
Dawkins,
Richard (1982). The Extended Phenotype,
Oxford University Press.
Dennett, Daniel
(2011). “Homunculi rule: Reflections on Darwinian populations and natural
selection by Peter Godfrey Smith”, Biology and Philosophy 26 (4): 475-488.
Godfrey-Smith,
Peter (2012). Frédéric Bouchard and Philippe Huneman (eds.) “Darwinian
Individuals” in From Groups to
Individuals: Perspectives on Biological Associations and Emerging Individuality,
MIT Press. Accessed from:
<http://www.petergodfreysmith.com/PGS_Darwinian_Individuals.pdf>
Macfarlane,
Robert (2016). “The Secrets of the Wood Wide Web”, The New Yorker, August Issue:
<http://www.newyorker.com/tech/elements/the-secrets-of-the-wood-wide-web>
Uncited:
Wilson, Robert
A. and Barker, Matthew, "The Biological Notion of Individual", The
Stanford Encyclopedia of Philosophy (Spring 2017 Edition), Edward N.
Zalta (ed.):
<https://plato.stanford.edu/archives/spr2017/entries/biology-individual/>.
Now that you've read that essay, I'm just start talking about how what you just read relates to the very deepest issues in metaphysics. Here goes.
I think that the better-known ‘species’ problem in the philosophy of biology is highly analogous to this problem of ‘biological individuality’, and I think that the metaphysics of Structural Realism helps us to see these similarities more clearly. It is, however, very complicated to show this, so bear with me.
I think that the better-known ‘species’ problem in the philosophy of biology is highly analogous to this problem of ‘biological individuality’, and I think that the metaphysics of Structural Realism helps us to see these similarities more clearly. It is, however, very complicated to show this, so bear with me.
There’s a great Philip Kitcher quote in a 2012 book I haven’t read
called Preludes to Pragmatism: Toward a
Reconstruction of Philosophy (I found it in Adam Hochman’s reply to Neven
Sesardic on race, discussed in my recent post “Solving Race”) which nicely sums
up my stance on the issue: “There is a nondenumerable infinity of possible
accurate maps we could draw for our planet; the ones we draw, and the
boundaries they introduce, depend on our evolving purposes” [150]. What does
this mean? It means that the Platonic idea that the philosopher’s job is to “carve
nature at its joints” introduces a false teleology which is seriously
misleading. If you want an equally pithy slogan for the alternative
metaphysical view, try this dialethic aphorism: nature has infinite joints and no joints. What I mean by this is
that, although certain sets of joints will have help us increase our store of
information, knowledge and our ability to predict the future far better than
others, we can’t say where the joints are really
because there is no place where the joints are really because there was no designer and nature itself doesn’t
carve (agents carve). To focus
specifically on biology, what this means is this: there are no absolutely True Biological Categories, there is no Objective Truth about how we should
taxonomise dogs and wolves, whether marsupials count as mammals, whether those
skeletons found in Morocco were really homo
Sapiens or proto-homo Sapiens, or whatever. There is likewise no Objective Fact about whether honey bee
colonies or coral and their algae are really
one organism or two, or whether (per the Gaia Hypothesis) the earth is
really an organism or just a homeostatic system with feedback cycles and
some policing mechanisms, or whether there are really human races or just clinally varying ethnic groups. As Godfrey-Smith apparently likes to
say, ESSENTIALISM IS DEAD!
Along similar lines (and with much relevance to this ‘problem of
individuality’), we should also note, as David Hume did back in 1737, that our fundamental
intuitions about the persistence of
macroscopic objects and living things make no empirical sense. What is the
sense in which that percept of a tree your brain processed ten minutes ago was
a representation of the same tree a
percept of which your brain is processing now? Only that the time slice you’re ‘looking
at’ now evolved directly from the one ten minutes ago. What cannot be true, no
matter how we want to believe it, is that the two time slices are identical. Why can’t this be true? Because
the two time slices have loads of
different properties, even just according to the macroscopic or standard,
anthropic descriptions (there are perceptible differences in position of
leaves, in position of ants on trunk, on the specific birds nesting or roosting
or resting, you know that there has been some capillary action inside the tree
to transmit water, and so on and so forth).
So it’s a mistake to say that “the same tree persists through time”! The
two trees are not the same tree! They’re different trees! Much closer to home
are those famous questions of personal
identity over time. We surely want to say that the five-year-old time-slice
with ‘my name’ is identical with the set of human time slices typing these
words. But any two five-year-old human time-slices are going to have more
properties in common than this one does with that one to which we nevertheless ‘want
to say’ I am identical… And if so, how does it make any sense at all to say “I
am the same person as I was when I was five”? [Hume made these observations
several hundred years ago, and I defend his bundle theory here (though I don’t
endorse my rejection of ‘perdurantism’ here for reasons we’ll come to): http://writingsoftclaitken.blogspot.com.au/2017/05/persistence-and-personal-identity.html].
Now, later, I’ll explain how Structural Realism helps us make sense of how this
can all be true while still allowing us to say that trees and people and chairs
are perfectly real and also that they
do in fact perdure. The key point is that recognising all this relativity
does not at all mean we slide into some kind of weird kind of Idealism or start
babbling nonsense about ‘texts’ like some kind of Pomo ninny. To understand why
it doesn’t mean this means understanding Structural Realism – so that’s where
we’ll turn to now.
I only very recently ‘got’ Structural Realism. After reading the book Every Thing Must Go at the beginning of
2016, the thing that most confused me – though I felt I learnt a huge amount from
the book and was taken with a lot of it – was how Ladyman and Ross and the rest
of the crew could simultaneously reject the ‘levels’ metaphor, and maintain
that oxygen, nitrogen, trees, animals, markets and “prices” (yes, prices) were real
(even if non-fundamental and ‘second-order’). This seemed to me like a
contradiction. I also strongly shrinked from their Quinean-type claim that any
old entity used in a scientific theory ought to be regarded as real simply if it ‘pays its rent’ in contributing to the scientific success (so to
speak). This particularly irked me when it came to their discussions of
economics (and I generally just disliked the fact they kept talking about
economics because I am a Steve Keen fan and (as far as politics and economics
go) only read post-Keynesian economists, Stiglitz, Chomsky and Peter Turchin,
and consequently have been led to believe that the entire economics profession
should be radically reformed (incidentally, I also constantly got this really
right-wing vibe from the book, not only because of the early footnote where
they randomly slag off Marx and the repetitive references to mainstream
economics, but because of the constant aggression and belligerence (only
ameliorated by the fact that they used ‘she’ as the default pronoun)).
It was only a couple of months ago that I suddenly understood how Ladyman and Ross could happily and consistently reject the ‘levels’
metaphor and mereology while maintaining that people, chairs, table, cats,
lemurs, ants, bacteria, species, prices, markets, oxygen, nitrogen, sulfuric
acid, (and so on and so forth) are all “real patterns”. The trick was being reminded of Dennett's discussions of Conway's famous Game of Life cellular automaton. Here's the takeaway:
In some sense, everything is
quantum fields (or whatever). This simply has
to be true. Fundamental physics is fundamental in the sense that (we’re
pretty sure) it describes phenomena to the same level of accuracy in every
region of the observable universe. You don’t need what Ladyman and Ross call a “locator”,
or an “address” for fundamental physics; it’s fundamental because the laws of fundamental physics are universal laws, describing universal structures of reality. Hence,
in some sense, everything is quantum fields. So there’s one level of reality,
and it is that described by fundamental physics.
So what’s with all this other less general shit? Where does it fit in? How can you be allowed to say everything that isn't fundamental physics can nevertheless be real if you insist that there’s only one level of reality? The
answer is to think about The Game of Life. Out of simple patterns in The Game
of Life you see more complicated patterns ‘emerge’ – patterns which are stable
and persistent and which, if you track them, allow you to compress a lot of
information about the dynamics of the system. What is the analogy with the real world? Well those stable, information-compressing patterns in the Game of Life have a direct analogue: any entity that earns its keep in institutionally approved and predictively successful science basically has to be one of those stable, information-compressing patterns. So how do we decide what is real of the patterns in reality that aren't the structures directly described by fundamental physics? Well, any kind of ‘projectible’ – stable,
trackable – pattern is real. And how do we determine the projectible patterns?
Well, our heuristic is that any ‘entity’ that pays rent in contributing
indispensably to scientific theories that achieve significant empirical success
in making predictions is a real pattern. And how have we avoided multiplying the
levels of reality? Because even though these patterns have a life of their own,
they are still patterns in fundamental
physics. You, me, that bug, jellyfish, amoebas and prices are projectible
patterns in the fundamental structures of reality (as Ladyman and Ross say,
this view dispenses even the need for distinguishing between types and tokens,
between categories and instances; real types and real tokens are both just
projectible ‘patterns’ (and so the problem of ‘species’ and the problem of ‘individuality’
really become extremely similar problems)). So there is one level of reality,
and we are patterns in it. There it
is!
Hopefully, it’s possible to see why this metaphysics allows us to
defend a kind of ‘perdurantism’ against the extreme Humean bundle theory that I
defended just before (days before) I
had these insights. An individual human is a stable, persistent, projectible
pattern. You can confidently track them as stable patterns, described in
different contexts by different fields (economics, anthropology, psychology) but
more or less stable in terms of properties, throughout their worldlines. So
they are real patterns and an individual human is a really persistent pattern
(a real four-dimensional worm), even if the individual time slices (of the
four-dimensional worm) are not really identical.
Something like that seems true anyway. (Incidentally, I’m
appreciating Ladyman and Ross’ work even more on the second reading. Every Thing Must Go really is an
impressive book.)
[1] It
should be noted that most of them clearly require conjunction with another one
of the candidates to constitute any kind of non-circular criterion for
biological individuality.
