Notes On Science, Metaphysics, and Understanding - Lecture

Overview

• Speaker’s context: LMU Munich talk; shift in emphasis toward metaphysics while keeping core content on scientific understanding. It blends philosophical analysis with historical case studies from a book chapter on scientific understanding.
• Core aim: investigate what scientific understanding is and how it relates to metaphysics; use a historical case study (Newtonian gravitation) to illustrate the view.
• Structure of the talk:
  • Present various positions on what scientific understanding is (Hempel, Friedman, unification, causal models).
  • Introduce the author’s contextual theory of understanding.
  • Show how the theory applies in a historical case study (Newtonian gravity).
  • Discuss the relation between metaphysics and understanding, including seventeenth- and eighteenth-century debates.
  • Conclude with implications and limits of understanding.

What is scientific understanding? Historical background and motivations

• Erwin Schrödinger quote: science is a Western invention rooted in ancient Greece; key feature is the hypothesis that the display of nature can be understood (science as intelligible, not magical).
• Problem for understanding in early philosophy of science: philosophers didn’t focus on understanding as an objective feature of science until recently.
• Logical empiricism and Hempel (1950s–60s): understanding is not part of the vocabulary of logic; explanations have a psychological/pragmatic element (subjective, contextual, not an objective feature of science).
• Friedman (1974): argued for an objective link between explanation and understanding; wanted an objective notion of understanding, aligned with the unificationist view (explanation provides understanding by unifying phenomena).
• Contemporary models of explanation: pluralism of models (unification, causal/mechanical, etc.).
• The central question: what does explanation do for understanding, and how is understanding produced across different models?

The contextual theory of understanding (the author’s view)

• Core claim: understanding is not reducible to one model; it depends on context (discipline, historical period, available skills).
• The grammar of understanding: a phenomenon p is understood by a subject S in a context C if there is a theory t that is intelligible to S in C and that helps explain p via a model linking t to p.
• Key ideas:
  • Understanding is pragmatic and contextual, not a one-size-fits-all objective relation between theory and phenomenon.
  • The relation between theory and phenomenon is mediated by a model: the model connects theory to phenomena and is often an idealization/abstraction requiring judgment and skill.
  • Skills are central: the ability to construct useful models depends on training, exemplars, and tacit know-how; algorithms alone cannot capture this process.
  • Explanation requires intelligibility of the theory; intelligibility is not intrinsic to a theory but depends on the context and the observer’s skills.
• Core criterion for scientific understanding (two parts):
  • There exists an explanation of p that is based on an intelligible theory t and that conforms to basic epistemic values of empirical adequacy and internal consistency.
  • Intelligibility is evaluated by how well scientists in a given context can use t to reason about p, including qualitative understanding without exact calculations.
• Intelligibility: a context-dependent quality
  • A theory t is intelligible to a scientist in context C if the scientist can recognize qualitatively characteristic consequences of t without performing exact calculations.
  • Intelligibility tests are teachable: students test intelligibility by applying t to simple problems and reasoning qualitatively.
• Concrete example introduced: kinetic theory of gases
  • Basic idea: a gas is a collection of molecules in motion; qualitative reasoning using this picture yields laws like that decreasing volume at fixed amount increases pressure and/or temperature.
  • Simple model: gas molecules as elastic spheres; diatomic molecule model with five degrees of freedom (3 translational + 2 rotational) as a simple idealization.
  • How intelligibility supports explanation: the simple picture gives qualitative insight into phenomena (e.g., specific heat), and refinements (more complex models) explain more detailed phenomena.
• Summary of understanding in this framework
  • A phenomenon is scientifically understood if there is an intelligible theory t that can explain p while meeting empirical adequacy and internal consistency, mediated by an appropriate model built with domain-specific skills.
  • This view allows multiple valid pathways to understanding, depending on context and available tools.

Intelligibility and testing: a formal gloss

• Intelligibility definition (in author's terms):
  • The value attributed to a cluster of theoretical qualities that facilitate the use of the theory in a given context.
• Two formal criteria for understanding (reiterated for clarity):
  • Explanatory requirement: there exists an explanation of p based on an intelligible theory t.
  • Epistemic requirements: the explanation is empirically adequate and internally consistent with the theory.
• Testing intelligibility in practice:
  • A theory is intelligible if a scientist can recognize qualitatively characteristic consequences of t without exact calculations.
  • This can be tested by applying t to simple problems and reasoning qualitatively about outcomes.
• Connection to education and Kuhnian ideas:
  • Skills and exemplars in education shape what counts as intelligible; the notion of intelligibility evolves with the community of practice.

Models of explanation and understanding: overview of positions

• Hempel’s covering-law model: explanations provide understanding but the relation is a logical, deductive one; understanding is a subjective/psychological by-product, not an objective feature of science.
• Friedman's unificationist model: explanations provide understanding by unifying diverse phenomena under fewer independent laws; objective, not subjective.
• Causal/mechanical models (Wesley Salmon, etc.): understanding arises when explanations reveal the causal mechanisms producing phenomena; causally salient explanations enhance understanding.
• Pluralism: no single model is universally correct; different contexts favor different explanatory strategies.
• The contextual theory integrates these views by letting the explanatory mode be chosen based on context, skill, and intelligibility in use.

The relation between metaphysics and understanding

• Traditional view: metaphysics has a close tie to intelligibility (e.g., Descartes’ requirement that knowledge be self-evident or mechanistic explanations reflect underlying metaphysical commitments).
• The author’s critique: the relation is not a simple one-directional dependence; metaphysics can both aid and hinder understanding depending on context and success of the tools it provides.
• Metaphysics as a toolbox for intelligibility:
  • Metaphysical ideas can supply concepts, mechanisms, or explanatory strategies that make theories intelligible within a given practice.
  • Successful metaphysical tools become restricted by their empirical success and may be retained or discarded as theories evolve.
• Historical illustration: Newton vs. Cartesian program
  • Huygens (Cartesian): rejected action at a distance because it seemed unintelligible within Cartesian metaphysics; sought intelligible causal mechanisms consistent with that metaphysics.
  • Newton: initially influenced by Cartesian ideas but diverged; he argued against an occult or inherent “activity” in matter as a cause for gravity; he entertained metaphysical explanations (e.g., God as mediator) but also emphasized the empirical/mathematical structure of gravitation.
  • Newton’s Principia: presented an inverse-square law and centripetal force; he explicitly refrained from speculating a mechanistic cause within matter and later entertained a metaphysical cause (theology) for gravity; he invoked “hypotheses non fingo” when it comes to gravity’s ultimate cause.
• Diagrammatic relation (descriptive only): Metaphysical ideas can provide tools for intelligibility (bridging theory and phenomena); successful tools reinforce those metaphysical commitments; failure or limited success can undermine them.
• The Cartesian toolset and Newtonian development illustrate how metaphysics can be both enabling and limiting depending on context and historical time.
• 18th-century shift and the oscillation of intelligibility:
  • Action at a distance moved from being problematic to becoming an intelligible, even foundational, tool in physics (Kant, Laplace, Helmholtz; Coulomb’s law).
  • In the 19th/early 20th centuries, new theories (e.g., Einstein’s general relativity) offered new intelligibility criteria, sometimes revisiting earlier concerns.
• Two competing positivist interpretations criticized by the author:
  • Mach/Frank: intelligibility is a hindrance that blocks progress; once new concepts arise, old intelligibility concepts fade.
  • Cushing: action at a distance was a failed attempt at intelligible discourse that was bracketed for centuries; later, Einstein’s theory restored intelligibility via a different framework.
• The author’s position: intelligibility is contextual and dynamic; metaphysics can provide useful tools but must be evaluated against empirical success and updated in light of new theories.

Newtonian gravitation: a detailed historical case study

• The basic claim: Newton’s gravitation was intelligible and productive only after negotiating metaphysical commitments about causation and distance.
• The Cartesian problem (17th century): action at a distance suggested occult powers and a magical universe; Cartesians favored contact (mechanical) explanations without mediation by distance or non-material causes.
• Huygens as Cartesian advocate for intelligibility via mechanism
  • Sought an intelligible cause of gravity that fit Cartesian mechanics; rejected action at a distance as unintelligible because it violated the mechanical picture.
• Newton’s position and evolution:
  • Early admiration for Cartesian program but exposure to alchemy, magnetism, and other phenomena that challenged Cartesianism.
  • In 1666, he derived an inverse-square relation for motion, a key mathematical insight, but did not yet articulate a full gravitational theory.
  • In the Principia (1687), he presented the law of gravitation and the centripetal force concept, along with inertia and other mechanics; he explicitly stated that he did not claim to explain the cause of gravity within matter.
  • The famous Newton quote: it is inconceivable that inanimate matter could act upon other matter at a distance without mediation by something non-material; this reflects his reluctance to commit to a Cartesian-like mechanical cause.
  • In the afterword and later writings (Scholium), Newton discusses potential metaphysical causes and the role of God as mediator; he treats gravity as a phenomenological law with a metaphysical backdrop rather than a purely mechanical explanation.
• Distinction between intelligibility and metaphysical intelligibility:
  • Newton and Cartesian thinkers used the term intelligibility in the sense of alignment with a metaphysical framework, which is distinct from the author’s notion of scientific intelligibility (theoretical qualities that facilitate use).
  • The shift from Cartesian to Newtonian metaphysics illustrates how intelligibility criteria can change with theory and context.
• 18th-century consolidation and broader acceptance of action at a distance
  • Figures like Kant, Laplace, and Helmholtz embraced action at a distance as intelligible and productive, expanding it to electricity (Coulomb’s law as another inverse-square law).
  • The move shows how historical context changes what counts as intelligible in science, aligning with the contextual theory of understanding.
• Positivist and anti-positivist interpretations contrasted
  • Positivists (Mach, Frank) saw intelligibility as a hindrance to progress that would be outgrown with new concepts.
  • Cushing argued that action at a distance was a failed attempt to achieve intelligibility and should be bracketed; Fourier/Einstein later offered new paradigms that altered intelligibility criteria again.
• Overall lesson from the Newtonian case
  • Metaphysical ideas can provide powerful tools for intelligibility and model-building, but their success depends on empirical adequacy and internal consistency of the theories they support.
  • Intelligibility is historical and contextual; shifts in metaphysical frameworks can reshape what scientists find understandable.
  • The oscillation of intelligibility across history demonstrates that there is no single, universal standard—understanding is contextual and historically contingent.

Implications: limits and scope of scientific understanding

• Limits of understanding are context-bound: understanding is always situated within a particular theoretical, methodological, and metaphysical context.
• The context can change, reconfiguring what counts as intelligible and what counts as an adequate explanation.
• Metaphysics as a toolset: not all metaphysical ideas are scientifically useful, but some provide concrete tools for constructing models and explanations.
• A non-dogmatic stance: there is no single path to understanding; different domains and