Transcript KaplanResponse

First, I’d like to thank the panelists for their thoughtful
comments on my book; I can’t tell you how honored I am
that so many people took the time to think seriously about
my work and respond to it. Also, I’d like to thank Sharyn
Clough for organizing this event; without her hard work
none of this would have happened.
In some ways I suppose that I should be pleased that there
were no serious criticisms of either the analysis of the
nature of selection or of the analysis of the level of
selection debates that Massimo and I presented. I bring
this up here because our interpretation of selection and
fitness (based on Matthen and Ariew’s work) informs the
approach we took to several areas that the panelists did
touch on.
So let me spend a few minutes of my time explaining, in the
broadest outlines, what we were up to.
Three distinct levels of analysis in evolution:
1) Individual
2) Population
3) Ensemble
The individual level focuses, roughly, on
individual organisms. More specifically,
it focuses on actual physical/causal
processes that interact with particular
entities capable of differential
reproductive success.
The ensemble level considers collections
of such populations, and as such is
essentially a tool for analyzing the
statistical distribution of the results of
processes at the individual level that effect
the population level.
Frequency
The population level focuses, roughly, on
actual populations – groups of entities
that experience relevantly sets of
physical/causal processes of the above
sorts.
Ratio of Growth Rates
Two Versions of Selection & Fitness
“Informal” or “vernacular” fitness:
An organism’s “propensity” to leave (successful) offspring in a particular
context (either overall in a particular environment, or given a particular kind of
interaction)
“Predictive” or “formal” fitness:
The average rate of increase of organisms in a population with one variant
of a trait as opposed to another variant of that trait.
Nature Selection in the informal
Sense:
The relationship of particular kinds
of physical processes and particular
variations in traits where that
variation is related by those
physical processes to differences in
reproductive success of some entity
Nature Selection in the formal
Sense:
The expected differences in
reproductive success of members of
a particular population divided on
the basis of the trait of interest.
So: Why does this matter?
I think it matters to at least the following issues Gene Centricism and Development:
Paul Roberts v. Karola Stotz
Microbiotic communities, “altruism,” and selection:
John Dupré
Functions versus causes: Paul Griffiths
And of course Metaphors, models, and conceptual rigor: Seth White
“Genes are (sort of) like a recipe. You of course
can’t get a cake without an oven, flour, sugar,
etc. But nonetheless the recipe has a certain
primacy, because it contains the information...
[G]enes are important, and different from other
components of the developmental process. With
all due respect to cytoplasmic inheritance and
maternal effects, there doesn’t seem to be much
here.” – Günter Wagner, in response to our book
The argument (?): Only heritable changes matter to evolution, and
since genes (as nucleic acid sequences) are the (only or only serious)
units of heredity, the only changes in populations that have
evolutionary consequences are those that involve changes in genes or
gene frequencies etc.
Note that this claim is independent of questions about e.g.
what gets selected in natural selection, etc.
Developmental Resources:
Resources necessary for development
Clinal variation in yarrowplant populations
(Achillea lanulosa), Sierra Nevada, California.
versus
Sources of heritable variation
What matters to biological evolution is heritable variation; if the
variation is heritable then it can have evolutionary consequences.
So evolution isn’t necessarily about changes in genes, but changes in
any heritable developmental resources at all. (Or, with respect to natural
selection and adaptations, at the very least any heritable variation that
can have fitness consequences…)
At the informal level, our attention should be drawn to physical interactions with
particular kinds of traits, and the ways that those traits are reproduced through
development
At the formal level, our attention should be focused on the
changes in trait frequencies, and what changes in developmental
resources those changes are associated with…
Non-genetic (or not-fully genetic) heritable developmental resources:
•
“Ordinary” epigenetic stuff: DNA methylation, chromatin condensation, etc.
•
Membrane Inheritance / Templating: Cortical membrane-based inheritance,
templating of other cellular structures including cytoskeletal structures.
•
Other intra-cellular stuff: Various material resources (including, in yeast at
least, prions), metabolic cycles and states, chemical gradients, etc.
•
Intercellular organization (in multi-cellular organisms): Tissue fields,
morphogen gradients, etc. (cellular inheritance through development)
•
Symbiotic inheritance: Bacteria, yeasts, etc. (including perhaps organelles such
as chloroplasts, mitochondria, etc.)
•
“Niche” construction and inheritance
• Environmental construction (figurative or literal)
• Environmental selection (passive or active) including host selection, etc.
•
Behavioral inheritance / learning: various forms, including at least passive
learning (via e.g. transmission of chemical signals re: food preferences),
differential attention to environmental factors, and/or active “mimicry.”
Few people doubt the importance of (many of) these resources for
development (in different cases), but are they important to
evolution?
Questions:
1) How common is the heritable variation in each (kind of)
system?
• How much of that variation is usually “visible” (to
selection, etc)?
• How much of that variation is usually suppressed?
• Under what conditions is that variation “released”?
2) How important is the heritable variation in each (kind of)
system? (and important for what?)
• In particular lineages?
• During particular (evolutionary) events or timeperiods?
If important for (key) evolutionary innovations,
then the heritable variation is important!
So even if, at some particular time, in some particular model organisms,
with some particular set of techniques, it looks like genetic variation “is
different from other components of the developmental process” and that,
with “respect to cytoplasmic inheritance and maternal effects, there doesn’t
seem to be much here” that is not yet a good reason to privilege genetic
over other developmental resources with potential heritable variation…
Developmental Niche Construction
Karola Stotz suggests extending our analysis to include
“Developmental Niche Construction” – a perspective which
“combines ideas of the active organism altering its environment
(niche construction), developmental systems theory and extended
(non- or extra-genetic) inheritance, evo-eco-devo and phenotypic
plasticity.”
With Karola, Massimo and I agree that understanding
evolutionary change, and perhaps especially innovations and
novelties, will require new ways of thinking about the genesis
and maintenance of stability as well as what happens when
stability breaks down, and I think we are each working
towards some version of the integrated perspective that
Karola recommends.
But while such a perspective will, I agree, be necessary for evolutionary biology to
fulfill its promise of explaining the origin, spread, and maintenance of traits in
populations, Massimo and I were (we thought) relatively cautious in the book to
permit room for such a perspective while not demanding one.
John Dupré draws out attention to microbes, and I’d like to link
some of those comments to the suggestions made by Karola.
For example, the “problem” of multi-cellularity is usually understood to be about how
cells learn to get along, and the answer is generally assumed to have something to
do with the genetic similarity of the cells in question. But this won’t do at all.
It seems increasingly plausible that many bacteria “species” can only live in close
symbiosis with other “species” of bacteria, and that these biofilms are most plausibly
thought of as multicellular organisms in their own right, complete with functional
specialization...
Now, it is true that there is extensive gene exchange in and among biofilm
communities, and some have suggested that it is this that maintains sociality in
bacterial communities. But there has been little attention to the possibility that
shared non-genetic resources are part of what maintain sociality. Biofilm
communities are clearly structured, and “altruism” can be associated with any
shared developmental resources, not just genetic resources…
Note that in structured hierarchical communities, development can (and
does) occur over different time and space scales, and each is important…
Paul Griffiths focuses on Massimo’s and my defense of a ‘modern history’
version of etiological functions, including in our individuation of one kind of
(or one notion of a) gene; Paul urges us to accept in addition causal
functions in at least some contexts, namely those linked to “causal
capacities of the organism that are relevant to understanding its
evolutionary fate.”
I’m going to resist that suggestion…
“Biting the bullet”
Taken seriously, the claim is that UCG has the function to code for Serine is
indeed hostage to discoveries about whether its place in that casual chain is
the result of selection or ‘merely’ of chemistry.
The phrase “codes for” however is without functional overtones. So
sure, UCG codes for Serin (notwithstanding concerns about the
informational ascription of coding talk). But that doesn’t mean that the
function of UCG is to so-code.
So what about the regress?
1) Ascriptions of selected functions are generated by (hypothetical) causal
analysis of the capacities of ancestral organisms to survive and reproduce
in ancestral environments (Griffiths 2003)
2) Hence, if we cannot identify which capacities of ancestral organisms to
subject to causal analysis without knowing what parts of those organisms
were selected for in their environments, then we face a vicious regress.
3) Therefore, a purely causal analysis of the adaptive role played by parts of
ancestral organisms must be possible without knowing what those parts
were adaptations for.
4) Furthermore, ancestral organisms cannot be easier to causally analyze
than living organisms on which we can actually experiment (Stotz and
Griffiths 2002)
Here I think keeping separate the formal and informal selection helps us
see why 1 through 3 don’t imply that there are non-etiological functions in
biology…
1) Ascriptions of selected functions are generated by (hypothetical) causal
analysis of the capacities of ancestral organisms to survive and reproduce
in ancestral environments (Griffiths 2003)
Yes – this is done by looking for opportunities for informal selection with
respect the causal capacities of ancestral organisms.
2) Hence, if we cannot identify which capacities of ancestral organisms to subject
to causal analysis without knowing what parts of those organisms were selected
for in their environments, then we face a vicious regress.
Here I want to note that informal selection of the sort appealed to in (1) doesn’t
require that formal selection (a statistical trend) has taken place. Hence we can
identify cases of informal selection without knowing whether that selection has
had any effect on the evolution of the traits in question.
So we don’t need to know what the traits in question were for (or even that they
were for anything) to apply the first part of (1).
3) Therefore, a purely causal analysis of the adaptive role played by parts of
ancestral organisms must be possible without knowing what those parts
were adaptations for.
Not exactly… A purely causal analysis of the way in which the parts in question
interact with a variety of events/forces in the world must be possible (informal
selection) without knowing what the parts are for (or even if they are for
anything)
The question of ‘for-ness’ is a question that appeals to the
formal sense fitness and selection – ‘for-ness’ ties the informal
causal analysis to the formal expected changes in the
distribution of traits…
4) Furthermore, ancestral organisms cannot be easier to causally analyze than
living organisms on which we can actually experiment (Stotz and Griffiths
2002)
Yup. But sometimes the assumptions regarding the likely course of the
evolution of a population – the likely links between the causal capacities of
the parts in question and the statistical outcomes – are likely to be correct…
Metaphors and science
Seth White notes that one of the main themes in the book is troublesome
metaphors. I should say that the working subtitle for the book was
“Metaphors and models in organismal biology.”
With Seth, I am confident that metaphors in science are not always
problematic, and I agree that as long as one remains focused on the
limitations of the metaphors currently in use, they can in fact be helpful.
I think as well that Seth’s interest in metaphors in ecology is well-placed;
there is a lot of work to be done unpacking and criticizing both contemporary
and past metaphors.
Notions like “invasive” species, “restoration” ecology, and the like are, I think,
ripe for serious critical work…
I could say more about this, but given the time limitations…