Contribution
A survey that argues for a reorientation. It collects in one place the reasons microbes matter to philosophy, from biogeochemistry to philosophy of mind. For each area it names the debate a microbial perspective changes: species concepts, the tree of life, major transitions, model organisms, individuality and causation. The authors’ own earlier work (O’Malley & Parke 2018, on comparability across model media) underpins §4.1’s claim that microbial experiments, mathematics, simulations and thought experiments can be compared.
Key insight
Philosophy of biology’s “zoocentric perspective” was a contingent inheritance from zoologist founders such as Mayr and Lewontin, not a principled choice (§1). Many phenomena thought distinctively animal (predation, ageing, kin recognition, even aspects of cognition) have microbial analogues that are easier to study experimentally. Microbes are therefore not an exception to general theory: they are often the best place to test it.
Assumptions
- The entry has no formal premises. Its authorities are microbiological and evolutionary literature (Woese, Doolittle, Lenski, Gause) and philosophy of modelling (Weisberg 2013; Levy & Currie 2015).
- No hard line divides microbes from non-microbes. “Microorganism” and “macrobe” have “no classificatory or evolutionary basis”, and “prokaryote” “obscures major evolutionary divergence of Bacteria and Archaea” (§1, Table 1).
- Models may be compared across media: mathematical, computational, material and mental (§4.1, citing O’Malley & Parke 2018).
Key results
What it argues:
- §1. The neglect of microbes was a matter of habit, not bias. Attention grew from the early 2000s: classification and phylogeny first, then individuality (Dupré & O’Malley 2007; Ereshefsky & Pedroso 2013) and the origin of life.
- §2.1. Microbes “run” the biogeochemical cycles, and their high turnover may explain why cycling is rapid rather than a long-term sink. Oxygen comes from cyanobacteria, so “philosophy is possible because of bacterial photosynthesis”. Symbiosis is pervasive (lichens, Wolbachia, the human gut), and “addiction” may describe hosts better than “benefit” (Hammer 2024). Most of life’s 3.8-billion-year history has been exclusively microbial.
- §2.2.1. Bacteria and Archaea defeat the biological species concept (no sex, little morphology, lateral gene transfer), which pushes philosophers to pluralism. The entry notes one dissent: Bobay & Gogarten 2017 say the BSC fits microbes when gene flow is tracked genomically.
- §2.2.2. Lateral gene transfer violates the vertical-descent-only graph of a tree, making the tree of life “a questionable construct… both metaphysically and pragmatically”.
- §2.2.3. The individuality framing of major transitions may not fit eukaryogenesis (Asgard archaea, late mitochondrial acquisition). Metazoan “signature” genes pre-date metazoa in choanoflagellates, filastereans and ichthyosporeans.
- §3. Pure culture, Koch’s postulates and germ theory turned medicine experimental. Molecular biology’s founding results came from E. coli and phages (Avery 1944; Hershey & Chase 1952). The operon model shows both how far microbial models generalise and where they stop.
- §4.1. Gause’s paramecia confirmed competitive exclusion but failed to produce Lotka–Volterra oscillations. Those were achieved experimentally only recently (Blasius et al. 2020, algae and a microscopic metazoan predator). Lenski’s long-term evolution experiment (12 populations since 1988, more than 82,000 generations by 2025) produced citrate use at about 31,500 generations. Frozen fossil records let populations be “replayed”, which realises part of Gould’s “tape of life” experiment. Avida runs the same logic in silico.
- §4.2. Microbial predation, ageing, kin recognition and cognition are productive models. Not identical to the animal phenomena, but indirect representations, as modelling generally is.
- §5.1. Individuality is approached through evolutionary criteria (“one Darwinian unit?”), not mereology. It is unresolved, and the ecosystem and community framings are live alternatives. Self and human-nature claims are vague. The authors separate three readings of them.
- §5.2. Causal claims about microbiomes are tested by philosophical accounts of causation, and in turn test those accounts. Confusing causal-role with selected-effect function slides into claims of unified individuality. Racialised microbiome research repeats the old problems of racial categorisation.
- §6. Microbiology bears on identity, selfhood, human nature and mind, not only on philosophy of science. The question of why gut microbes, but not helminths, are readily counted as part of the body is left as an exercise.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Philosophy of biology’s neglect of microbes was contingent habit, not active bias | weak | assertion with a historical sketch (§1, n. 1) |
| C2 | No single species concept works across microbial diversity, and philosophers default to pluralism | moderate | secondary literature (Doolittle 1999; Ereshefsky 2010; Dupré 2001), with the Bobay & Gogarten dissent reported |
| C3 | Lateral gene transfer makes the tree of life a questionable construct | moderate | cited molecular phylogenetics (Doolittle & Bapteste 2007; Dagan & Martin 2009) |
| C4 | The origin of eukaryotes may be mainly diversification within Archaea, not a transition in individuality | weak–moderate | recent phylogenomics (Williams et al. 2020; Vosseberg et al. 2024; Williamson et al. 2025), flagged as contested (Bremer et al. 2022) |
| C5 | Microbial systems can be compared directly with mathematical, computational and even mental models | moderate | case studies (Gause, Lenski, Avida) and the authors’ earlier argument (O’Malley & Parke 2018) |
| C6 | Predation, ageing, kin recognition and cognition have productive microbial models | moderate | cited reviews (Pérez et al. 2016; Baig et al. 2014; Wall 2016; Beekman & Latty 2015) |
| C7 | Whether host and microbiome form one unit of selection is unresolved | moderate | report of a live literature (Booth 2014; Godfrey-Smith 2015; Pradeu 2016; Skillings 2016) |
| C8 | Claims that the microbiome changes the “self” are usually vague, and there is no evidence that microbes affect first-person experience | moderate | conceptual disambiguation (§5.1), citing Parke et al. 2018 |
| C9 | Racialised microbiome research perpetuates the problems of racial categorisation | moderate | cited critiques (De Wolfe et al. 2021; Delgado & Baedke 2021) |
Concepts
- microbe / microorganism — informally, a unicellular organism, or anything viewed microscopically. It has no classificatory or evolutionary basis (Table 1). The entry’s working sense is organisms able to live and reproduce as a single cell “most (but not necessarily all) of the time” (§1, n. 3).
- zoocentric perspective — treating animals, especially humans, as the paradigm organisms (O’Malley & Dupré 2007).
- frozen fossil record — revivable frozen samples of ancestral populations, which allow evolutionary trajectories to be reconstructed and rerun (Lenski 2011).
- individuality (microbiome sense) — “individuating where one biological entity starts and another begins”. Here it is answered by evolutionary criteria (is it one Darwinian unit, one unit of selection?) rather than by mereology.
- causal-role vs selected-effect function — what a microbe does, versus what selection has made it for. Conflating them turns causal findings into claims that host and microbiome are unified (§5.2).
Connections
For its individuality material the entry points to the SEP entries on Biological Individuals (LIT-168) and Units and Levels of Selection (LIT-160). Its §5.1 question, whether host and microbes are one Darwinian unit or unit of selection, is the one LIT-160 §4.6 argues at length. There Lloyd defends holobionts as interactors and cites O’Malley’s work with Dupré. LIT-168 places squid+Vibrio as an organism and living agent that is not a Darwinian individual. That matches Lloyd’s report (LIT-160 §4.6) that Godfrey-Smith’s view excludes the squid from the units of selection. The ecosystem framing this entry reports (Skillings 2016; Schneider 2021) is a third option, between counting such pairs as one individual and counting them as two.
Baedke’s The Organism (LIT-167) argues that holobionts meet the “persister” criterion too. That fits §5.1’s report that host–microbe individuality is unsettled. Baedke is cited here only for other work (Delgado & Baedke 2021 on race and microbiomes).
Griesemer & Shavit (LIT-188) and Bourrat (LIT-157, LIT-049, LIT-005) work on the ETI criteria of individuality that §2.2.3 puts in question for eukaryogenesis. The entry cites none of them.
Meincke’s process view of pregnancy (LIT-142) shares the question of whether one organism contains another, but is not cited.
The account of identity it presupposes: individuality is biological and evolutionary, not metaphysical. Individuating a body means asking whether its parts share an evolutionary fate. The entry is agnostic about the answer and deflationary about selfhood. It treats “self” talk as ambiguous among biology, composition and phenomenology.
Bearing on the record
Nothing for ML practice. The one potentially transferable idea is §4.1’s comparison of material, computational and mental models, including Lenski’s “run like algorithms” experiments against Avida. That is a point in philosophy of modelling, not an instruction.
For the record, the entry supports treating biological individuality as open. Specifically, it supports treating individuality as the organising theme of major transitions as open. Any document summarising “major transitions = transitions in individuality” as settled should note the Asgard counterpoint the entry reports. The dossier’s link to the reciprocal-causation works (LIT-110, LIT-156) should be dropped or downgraded to “causation in biology”.
Limitations
- It is an orientation, not an argument for a thesis. Every philosophical position is reported and attributed. The authors’ own commitments are the programme of attention itself and the deflation of microbiome-and-self claims.
- Individuality gets two paragraphs (§5.1). Readers need LIT-168 and LIT-160 for the substance.
- The Asgard claim rests on phylogenetic inferences that the entry itself calls contested and “still to be divulged”.
- The bibliography gives Maynard Smith & Szathmáry as 1997 (the OUP paperback). The original is 1995. This is harmless, but a date citing the entry may copy.
Open questions
- When does a host–microbe association become one evolutionary unit, and which criteria decide it (§5.1)?
- Whether eukaryogenesis was a transition in individuality or a diversification within Archaea. To be settled by further phylogenomics of Asgard relatives (§2.2.3).
- How far material microbial models can stand in for mental or thought-experimental models (§4.1, “just how far such comparisons can be taken”).
- How microbiome “function” should be attributed, whether as causal role or selected effect, and whether the patient to be diagnosed is a multispecies entity (§5.2, Inkpen 2024).
Corrections to the seeded skim
- The dossier (§2.2.3) says major transitions “look different from a microbial vantage”. The full text makes a sharper claim, new in this revision. Asgard-archaea inferences (Williams et al. 2020; Vosseberg et al. 2024; Williamson et al. 2025) place eukaryotes inside Archaea, with the mitochondrial endosymbiosis probably “late”. So the origin of eukaryotes “might not be driven primarily by a transformation in individuality (as most philosophers think it is)”. The entry calls this contested (Bremer et al. 2022). It also flags the individuality framing of major transitions as itself disputed (“however, cf. O’Malley and Powell 2016”). This is the entry’s most direct bearing on the identity tag, and the dossier does not record it.
- The dossier’s summary says taking microbes seriously “changes debates over… biological individuality and causal explanation”. The entry is more cautious than that. On individuality it reports that criteria for when symbionts become one evolutionary unit “are still disputed”. It adds that “for some commentators” an ecosystem or community framing explains more (§5.1). It opens §5 warning against “exaggerated views” of microbiome effects on mind and health (Hooks et al. 2019; Mitchell et al. 2026). It separates three readings of “microbiomes change the self”: controversial biological individuality, a factual point about bodily composition, and a phenomenological claim. On the third it says “to date, there is no evidence that microbes influence this” (§5.1).
- The dossier suggests §5.2 links to the reciprocal-causation debate (LIT-110, LIT-156). The entry never mentions reciprocal causation. §5.2 is about justifying causal claims in microbiome science: a “causal law” account (Gillies 2019), refinements to interventionism (Attah et al. 2020; Oftedal 2020), and how determinative microbiome causes are (Suárez 2024). It also covers causal-role vs selected-effect function, and the risk of racialised microbiome categories. The link to LIT-110 and LIT-156 is only that both concern causation in biology.
- Omitted by the dossier: §2.1.1 frames host–microbe symbiosis as possibly “addiction” rather than benefit (Hammer 2024) and discusses the Gaia hypothesis. §5.2 criticises microbiome studies that racialise microbiomes, inverting old rankings without escaping racial categorisation (n. 8). §4.2’s list of phenomena once defined by animal traits includes cognition as well as predation, ageing and kin recognition.
- The dossier’s date: 2020 is first publication. The version read is the 10 April 2026 substantive revision, which cites work up to 2026 (e.g., Mitchell et al. 2026; Williamson et al. 2025; Halperin 2025). It reports the Lenski experiment at “more than 82,000 generations” in 2025, now in Barrick’s laboratory (n. 7).
- identity tag: justified. §5.1 is about individuation (“where one biological entity starts and another begins”) and about the “self”. §2.2.3 is about individuality in major transitions. §6 names “identity, selfhood, human nature” as topics microbiology bears on. The tag reflects a real part of the entry, though a minority of it.
- primary topic: philosophy-of-science (keep).