How Mechanisms Replaced Laws as the Core of Explanation in Biology and Life Sciences
▶️ Rave the World Radio
24/7 electronic music streaming from around the globe
William Bechtel is a philosopher of science at the University of California, San Diego (UCSD), where he is a professor in the Department of Philosophy and affiliated with the Science Studies Program. He is a prominent figure in the "new mechanist" movement in philosophy of science, particularly focusing on how explanations work in the life sciences (such as biology, neuroscience, and cognitive science).
What is the Mechanistic Approach to Explanation?
Bechtel advocates for mechanistic explanation, which means that to explain a scientific phenomenon (e.g., how a cell metabolizes sugar, how memory is formed, or how a biological process occurs), scientists identify and describe the underlying mechanism responsible for it.
A mechanism, in this context, is typically understood as:
- A system of component parts (e.g., molecules, cells, neurons, or sub-systems).
- Performing specific operations or activities.
- Organized in a particular way (spatially, temporally, or functionally) so that their interactions produce the phenomenon in question.
For example, explaining cellular respiration mechanistically involves detailing the parts (like enzymes, mitochondria, electron transport chain components), their operations (oxidation-reduction reactions, proton pumping), and how they are arranged to generate ATP from glucose. Scientists often use diagrams, models, simulations, and multilevel descriptions (from molecular to organism level) to represent these mechanisms.
This approach is causal and decompositional: it breaks down complex behaviors into how parts work together, often iteratively refining the model as new evidence emerges.
Contrast with the Traditional Model: Deductive-Nomological (D-N) Explanation
Mainstream philosophy of science in the mid-20th century (influenced by figures like Carl Hempel) promoted the deductive-nomological (D-N) model (also called the covering-law model). According to this view:
- A good scientific explanation involves deriving (deducing) the phenomenon (the explanandum) logically from:
- General laws (universal statements, often expressed mathematically).
- Specific initial conditions.
- The explanation takes the form of a logical argument: If the laws and conditions are true, the phenomenon must follow deductively.
For instance, explaining why a pendulum swings at a certain rate would involve deducing it from Newton's laws of motion plus initial conditions like length and gravity.
Bechtel argues that this D-N framework does not accurately capture how actual scientific practice works, especially in the life sciences. Biology and related fields rarely rely on strict, exceptionless universal laws like those in physics. Instead:
- Life scientists explain by identifying mechanisms — describing entities, activities, and organization.
- Explanations are often multilevel, dynamic, and involve diagrams, simulations, and models rather than purely linguistic/logical deductions from laws.
- Phenomena are accounted for by showing how parts interact to produce the behavior, not just by subsuming it under a law.
- Mechanistic explanations allow for cascades of explanations (e.g., a higher-level mechanism explained by a lower-level one), and they better handle complexity, nonlinearity, and context-dependence common in biology.
In short, Bechtel contends that the mechanistic approach better aligns with the actual methodology and explanatory practices of biologists and life scientists. It offers a more realistic and productive alternative to the traditional law-based, deductive framework that dominated earlier philosophy of science.
His influential works, such as Discovering Complexity (with Robert Richardson) and the paper "Explanation: A Mechanist Alternative" (with Adele Abrahamsen), elaborate on these ideas and helped shift philosophical attention toward mechanisms as central to understanding scientific explanation in complex domains.
Comments
Post a Comment