I confess I am a bit bewildered here. Did I miss a memo from Claude?
Nick
*From:*Friam <[email protected]> *On Behalf Of *glen
*Sent:* Thursday, August 7, 2025 3:39 PM
*To:* [email protected]
*Subject:* Re: [FRIAM] Group Selection IS a metaphor.
Here is Perplexity's assessment of Claude's analysis of your attempt
to buttress group selection. Claude's analysis is path dependent (as always).
And a critical part of the prompt-pathway is my fascination with the claims
against evolution made by (even more ignorant than me) Intelligent Design
advocates (IDers). A significant problem with biologists' attempts to
*charitably* address issues raised by IDers is that the IDers are not humble
enough to learn what *working* evolutionary biologists actually *do*. Claude
was charitable to you, Nick, too in saying that your work is perhaps too
philosophical and/or metaphorical to be directly relevant to current work in
evolutionary biology. So my own conclusion (as incompetent as it is) might be
that your paper probably applies well as a case to study in a theory of
metaphor, but is largely irrelevant to evolutionary biology. That's OK. But as
Dave points out, we want to be clear about the domain we're in: a theory of
metaphor *or* hypothesis formulation in biology. I claim we shouldn't conflate
those two domains.
Anyway, here's Perplexity's assessment of Claude's claims:
1. Gardner (2015): Mathematical Resolution of Thompson’s
Ambiguities
*Claim:* Gardner’s work rigorously separates individual and group contributions
to evolutionary change, resolving Thompson’s “flock” ambiguity with the Price
equation.
*Is Citation (Gardner 2015) Appropriate?*
* Yes. Gardner (2015) explicitly develops a genetical theory of
multilevel selection and uses the Price equation to mathematically
decompose group vs. individual effects, addressing foundational issues
of operationalizing group selection.^1_1
<https://academic.oup.com/jeb/article/28/2/305/7381362>
*Improvement:* Citation is appropriate and robust for this claim.
2. Simon et al. (2013): General Mathematical Solution for
Multilevel Selection
*Claim:* Simon et al. (2013) unify multilevel population processes with
mathematical modeling, clarifying group “parents” via explicit group
reproduction and composition.
*Is Citation Appropriate?*
* Yes. Simon et al. (2013) address the need for explicit multilevel
event modeling and use continuous-time Markov chains to clearly
distinguish group- and individual-level events, directly tying into
both conceptual and mathematical ambiguities.^1_3
<https://pubmed.ncbi.nlm.nih.gov/23730751/>
3. Price (1970/1972): Foundational Price Equation
*Claim:* The Price equation provides a recursive partitioning formula for
selection at any level, allowing for formal group fitness definition.
*Is Citation Appropriate?*
* Entirely correct. Price’s original and follow-up work establish
the equation’s centrality. See Price (1970) and Price (1972), both
cited broadly as the mathematical basis for partitioning selection at
multiple levels.^1_5 <https://www.nature.com/articles/227520a0>
4. van Veelen et al. (2012, 2014): Mathematical Non-Equivalence of
Frameworks
*Claim:* van Veelen and collaborators demonstrate critical limitations, showing
group selection is not always reducible to inclusive fitness or the Price
equation.
*Citations:* van Veelen (2012), van Veelen (2014).^1_6
<https://pubmed.ncbi.nlm.nih.gov/21839750/>
* Correct and highly relevant: both papers provide modern mathematical
counter-examples and directly engage with the limits of equivalence between
approaches.
5. Traulsen and Nowak (2006): Stochastic Group Formation Model
*Claim:* Group structure can emerge naturally from population structure; offers
a mechanistic, not arbitrarily-imposed, solution for group selection modeling.
*Citation:* Traulsen & Nowak (2006).^1_8
<https://pubmed.ncbi.nlm.nih.gov/16829575/>
* Accurate: Their model lets groups arise dynamically rather than being
defined a priori, fulfilling the intended rhetorical point.
6. Hamilton (1964a,b) & Queller (2011): Inclusive Fitness Theory
and Extensions
*Claim:* Hamilton’s “rb>c” rule sidesteps group boundary issues. Queller
generalizes to “kin, kith, kind,” separating individual and social effects
quantitatively.
*Citations:* Hamilton (1964), Queller (2011).^1_9
<https://www.pnas.org/doi/10.1073/pnas.1100298108>^1_11
<https://www.sciencedirect.com/science/article/pii/0022519364900396>
* Both are strongly appropriate. Hamilton founded inclusive fitness
theory and Queller explicitly expands it to broader social
contexts.^1_10
<https://en.wikipedia.org/wiki/The_Genetical_Evolution_of_Social_Behav
iour>
7. Gardner (2015) and Structured Populations: Unresolvable Class
Structures
*Claim:* Gardner (2015) demonstrates that group decomposition is sometimes
impossible in highly structured populations, revealing new analytical tensions.
*Citation:* Gardner (2015).^1_2
<https://pubmed.ncbi.nlm.nih.gov/25475922/>
* Supported: Gardner discusses “irreconcilable classes” and the challenges
of meaningful within- and between-group covariance calculations in structured
settings.
8. Marshall (2011) vs. van Veelen et al. (2012): Ongoing Debate on
Framework Equivalence
*Claim:* Marshall argues for, van Veelen against, universal equivalence of
group selection/inclusive fitness (formal equivalence controversy).
*Citations:* Marshall (2011), van Veelen (2012).^1_12
<https://pubmed.ncbi.nlm.nih.gov/21839750/>^1_6
<http://web.evolbio.mpg.de/~garcia/preprints/08.pdf>
* Accurate and well-chosen citations. The Marshall review discusses
equivalence claims; van Veelen et al. provide mathematical dissent.
9. Fletcher & Zwick (2007), Nowak (2006), Hilbe et al. (2018):
Explicit Parameterization via Game Theory
*Claim:* Game theory resolves ambiguities by making group formation, structure,
and trait attribution explicit in the model.
*Citation:* Nowak (2006), Hilbe et al. (2018).^1_8
<https://pubmed.ncbi.nlm.nih.gov/16829575/>
* Nowak’s “five rules” for the evolution of cooperation are the field’s
standard reference for mechanistic, parameterized modeling.
* Hilbe et al. (2018) extends classical models for information flow and
flexible group boundaries; relevant as cited.
10. Wade (1976, 1977), Philson et al. (2025): Empirical Approaches
*Claim:* Wade’s experiments operationally define groups and fitness, Philson’s
fieldwork on wild marmots uses social network methods for group boundaries and
traits.
*Citation:* Wade (1976), Philson et al. (2025).^1_8
<https://pubmed.ncbi.nlm.nih.gov/16829575/>
* Citation to Wade is historically foundational and warranted.
* Philson et al. (2025) (if published as stated) would be highly relevant
for contemporary empirical methodology.
11. Wynne-Edwards (1962): Historically Important, Theoretically
Ambiguous
*Claim:* Wynne-Edwards raised group-level explanations, often ambiguous, but
some cases may be valid under modern theory.
*Citation:* Wynne-Edwards (1962).
* Correct for historical reference; less so for modern resolution, but that
is acknowledged in the content.
Conclusion: Citation Audit
* *Most citations are accurately and specifically applied to the claims
made.* No clear mismatches requiring replacement were detected for major
rhetorical points.
* Pivots like the Price equation, the debate on equivalence (Marshall/van
Veelen), and modern empirical evolution studies all use canonical and robust
literature.
*If a citation is missing or weak*—for example, if you wish to cite a specific
field study for social network-based group delineation and cannot access Philson et
al. (2025)—recent field network analyses in wild animals (see Farine &
Whitehead, 2015, “Constructing, conducting and interpreting animal social network
analysis”) would serve as an alternative.
Summary Table: Citation Appropriateness
*Theoretical Point*
*Citation(s)*
*Appropriateness/Relevance*
*Replacement Needed?*
Mathematical formalization/Price equation
Gardner (2015), Price (1970, 1972)
Excellent
No
General theory of multilevel processes
Simon et al. (2013)
Strong
No
Inclusive fitness and boundary issues
Hamilton (1964a,b), Queller (2011)
Field-standard
No
Equivalence/non-equivalence debate
Marshall (2011), van Veelen (2012, 2014)
Specifically relevant
No
Mechanistic/game-theoretic models
Nowak (2006), Hilbe et al. (2018)
Industry standard
No
Empirical/operational definition (beetle experiments, wild networks)
Wade (1976), Philson et al. (2025)
Foundational/contemporary
No (unless Philson unavailable)
Historical ambiguity/group adaptation
Wynne-Edwards (1962)
Canonical for controversy
No
*Recommendation:* Your bibliography is thoroughly curated and defensible for each
major rhetorical or theoretical point in the analysis. If more topical or current
empirical works are needed, specifically for fieldwork using social networks, cite
Farine & Whitehead (2015) as a supplement to the Philson reference if necessary.
On 8/5/25 11:10 AM, Nicholas Thompson wrote:
Dear Colleagues in FRIAM,
Sometimes, if I am going to get anything done, I just have to ignore
Friam, and keep my head down, and work at the thing I am working at. It always
seems, on that occasion, that you-guys dangle in front of me some enticing
topic so I must scream and put my fingers in my ears to keep focus on my work.
So it was that when I decided I must fish or cut bait on entropy or it would
take me to my grave, that almost immediately you-guys started not one but two
conversations close to my heart: on the centrality of metaphor to science and
on the group selection controversy.
A couple of decades ago I brought those two interests together in a paper
called “Shifting the Natural Selection Metaphor to the Group Level. There are
two things about this paper that make it salient for me. The first is that I
think it is the best paper I ever wrote. The second is that for each of the
two people whom I most hoped to reach when I wrote it, D. S. Wilson and Elliott
Sober, it is a piece of crap. In it, I try to show that the problem with
metaphors is not with their use in scientific thinking: on the contrary, it is
with their ill-disciplined use. Metaphors need to be worked in a systematic
way, not simply flung out in a gust of poetic exuberance. This lesson I try
to teach by working the natural selection metaphor in a systematic way to show
that if it had been treated seriously in the first place, the whole dispute
about group selection might have been avoided. Thus the paper is not only
arrogant, but meta-arrogant.
Nothing is more pitiable than the retired academic who would do anything
to have anybody read his moribund essays. But, alas, I simply am such a
person. So, I am attaching a copy of the paper in the hope that it will have
some value to you within the context of your two discussions.
Mumble,
Nick
