Ways of knowing

– a philosophical debate for microbial ecology

Aug 09, 2024

What (and how well) do we know of and understand the ecology of microorganisms? What are and how should we answer the grand scientific questions of the field? What constitutes progress in the discipline? The resolution of these questions of course has particularly strong parallels in particular to macrobe (sensu O’Malley and Dupre) ecology, but also fits within the broad field of philosophy of science.

Karl Popper’s ‘critical rationalism’ has had enormous influence on modern science. In Popper’s view, Science advances by posing ‘risky’ hypotheses that can be empirically tested to see if they are falsifiable. By risky, he meant either (a) content-rich, (b) unlikely according to current theory or (c) predictive of new phenomena. For Popper, proper science proceeds by deduction. By this standard, microbial ecology (and macrobe ecology) are failures. Prosser (2020) examined 100 published papers from five major microbial ecology journals and found that only 10 aimed to test a hypothesis. Betts et al. (2021) examined a broad swath of the ecology and evolution literature and estimated about 25% of papers presented hypotheses.

Popper’s views were a reaction against positivism and the long-standing views of Sir Francis Bacon on inductive reasoning in science. But Popper’s criterion for demarcation of science from non-science is quite restrictive and would rule out a great deal of accepted scientific practice. Imre Lakatos had a more moderate opinion -- assessment of scientific progress should be based on analysis of research programs, which are a series of theories linked via a set of fundamental assumptions. For Lakatos, a program is good science if it is progressive – either theoretically (by producing novel hypotheses) or empirically (if the novel predictions are empirically validated). A research program consists of a “hard core” of principles that everyone in the field accepts (Thomas Kuhn might call this the paradigm). Researchers augment this hard core with “auxiliary hypotheses” that are then empirically tested and subsequently altered or abandoned – if they are not falsified, they are retained if they enhance the program’s explanatory power.

Progress in Biology

Much of the work on philosophy of science has been done with an eye towards physics. But biology is quite a different animal (pun intended). Generalizations in biology never rise to the level of laws as found in physics – they might not hold depending on the historical context for example. In the realm of biology, progress is commonly viewed as the process of discovering mechanisms. Hence, mechanisms play the role of laws in biology. To the extent that research in both microbe and macrobe ecology is focused on demographics and detecting patterns, it is often described as ‘descriptive.’ What critics may be saying is the studies are rarely explanatory and do not provide insights into causation. In and of itself, such studies rarely lead to novel hypotheses or empirical tests of hypotheses.

The definitions of ‘mechanisms’ generally contain four elements:

a phenomenon
parts
causings
organization.

Represented in this way, a mechanism can be decomposed from whole-system behavior into the organized interactions among its active parts.

The meaning of causings is more contentious among philosophers of science than one might expect and has had multiple interpretations. To me, a useful frame is to think that a system part A causally impacts a part B if altering the value of A effects a change in B. Note that this need not mean there is direct physical interaction between the parts.

The meaning of ‘organization’ also has been broadly discussed by philosophers. It can have both spatial and temporal dimensions – the causings of parts within the organizational framework is what produces the emergent properties of the system. There are multiple levels of organization in any system. This should make intuitive sense to microbiologists – for example, in a biofilm the entire matrix is shaped by external forcing factors such as fluid flow and resource concentrations but this is mediated via the actions of individual cells each driven by molecular events that drive changes in gene expression.

Demographic analyses of microbial community composition and biogeochemical rate measurements in situ can identify phenomena and parts. With appropriate tools and consideration of the physico-chemical environment, spatial and temporal organization of habitats might be assessed. Analysis of the recent literature however indicates that microbial ecology is stuck primarily at description of phenomena and parts. Analysis of spatial or temporal organization at the true scale of microbes (the microenvironment) is quite rare, and little progress is being made on fundamental causes. My Web of Science interrogation of the Big 5* journals over 10 years found about 200 out of 11,000 publications in which the title included ‘mechanism’ or ‘cause’.

Nonetheless, ‘descriptive’ studies have a place in microbial ecology – when they identify important phenomena and parts. Metagenomic approaches might also be able to explain which organism(s) potentially are responsible for a process such as sulfate reduction or nitrogen fixation – with the caveat that their actual role cannot be determined by genomic approaches alone. The present problem in the literature is that there rarely is a follow up to the deeper (and more difficult) elucidation of how parts interact and cause the phenomenon of interest. Laboratories that are adept at the sequence of (1) isolate DNA from habitat, (2) sequence DNA, (3) perform exploratory or interpretive statistical analysis of binned sequences, and (4) speculate upon results most commonly move on to another habitat to perform a similar analysis, rather than digging more deeply into causal mechanisms.

This last failing is a serious one for the future of microbial ecology. Returning to Imre Lakatos’s view, a research program is only worth working on if it is progressive, and intellectually suspect if it is degenerating. Progressive programs are theoretically progressive (that is, their theories predict unexpected facts) and empirically progressive (some of the new “facts” must be experimentally corroborated). A program is degenerating if it is not delivering novel predictions, or the predictions cannot be experimentally corroborated. It is worth asking where the various strands of microbial ecology stand with respect to Lakatos’s standard.

I make these criticisms in full knowledge of how difficult the study of microbial ecology is. To quote Tom Brock (1987):

“Microbial ecology is not easier than those other things, but harder. Because microbial ecology deals with complex systems, it requires skill, insight, imagination, attendance to quantitative detail, even mathematical insight, that far surpasses what is needed for genetics and molecular biology.”

Your weekly moment of Zen …

Fulton Center, Manhattan

References

Betts MG et al. (2021) When are hypotheses useful in ecology and evolution? Ecology and Evolution 11: 5762-5776. https://doi.org/10.1002/ece3.7365

Brock, T.D. (1987) The study of microorganisms in situ progress and problems. in Symp. Soc. Gen. Microbiol., Volume 41. Ecology Of Microbial Communities

O’Malley MA and Dupré J (2007). Size doesn’t matter: towards a more inclusive philosophy of biology. Biology and Philosophy, 22(2): 155–191.

Prosser James I. 2020. Putting science back into microbial ecology: a question of approach Phil. Trans. R. Soc. B 375:20190240. http://doi.org/10.1098/rstb.2019.0240

Think Like a Microbe

Discussion about this post