Natural Sciences — Study Guide
Use this as a review reference for the Sciences major essay. The unit knowledge question is: “Is there solid justification for regarding knowledge in the natural sciences more highly than knowledge in another area of knowledge?”
Scope
What Makes a Natural Science?
The natural sciences study the natural world independent of human beings — their subject matter (hydrogen, gravity, cells) would exist even if humans did not. This is the defining feature.
- Method alone does not define the AoK — psychology uses experiments and economics uses models, but neither is a natural science. The key is what is studied, not how.
- The scientific method is designed to cancel out the human component: results must be replicable regardless of who performs the experiment or where.
- Science deals in brute facts — facts independent of any observer’s beliefs, values, or culture.
Key disciplines: Physics, Chemistry, Biology, Earth Sciences (each offering different aspects of the same natural world — water is H₂O in chemistry, a solvent in biology, and a fluid in physics).
“Science is a way of looking for natural explanations for all phenomena.” — Michael Shermer
Thinker to note: John Dupré (The Disorder of Things) — argues there is nothing special that marks out the natural sciences. A minority but important view.
Pure vs Applied Science
| Pure Science | Applied Science / Engineering | |
|---|---|---|
| Goal | Knowledge for its own sake | Solve a practical problem |
| Direction of fit | Knowledge changes to fit the world | The world changes to fit the specification |
| Level of generality | General theories, abstract models | Must account for messy local conditions |
Case study — Grindavik, Iceland (2023–24): Seismologists detected a large magma build-up beneath the town. They could not predict when an eruption would occur — but risk assessment led to evacuation. No lives lost. Applied science does not need predictive certainty to save lives.
Note: Many sciences are both pure and applied (ecology, seismology). The distinction is not hard and fast.
Science, Non-Science, and Pseudoscience
| Category | Definition |
|---|---|
| Science | Systematic, testable knowledge of the natural world |
| Non-science | Legitimate knowledge of things that are inherently human (arts, history, human sciences) |
| Pseudoscience | Claims that look scientific but cannot be properly tested |
Pseudoscience criteria: A claim is pseudoscientific if it uses a closed explanation — one that can explain any outcome and therefore cannot be falsified.
Examples: astrology (predictions too vague to test; stars in a “constellation” lie at vastly different distances), homeopathy (fails rigorous controlled testing; proponents invoke “negative energies” to explain failures).
Seven criteria distinguishing science from pseudoscience: replicable, consistent, observable, natural, predictable, testable, tentative.
Karl Popper — Falsificationism: A claim is scientific only if it is possible, in principle, to show it is false. Science should try to falsify hypotheses, not confirm them.
“In a real area of knowledge, researchers can (and sometimes do) get things wrong. In pseudoscience, being wrong is not an option.”
Methods & Tools
Types of Explanation
| Type | What it answers | Example |
|---|---|---|
| Constitutive | What is it made of? | Water is H₂O |
| Causal | Why did it happen? | Lightning is electrostatic discharge |
| Reductive | Explained in terms of smaller parts | Body temperature via molecular movement |
| Holistic | Explained in terms of the whole system | Gaia hypothesis (Lovelock) |
Occam’s Razor: Among competing explanations, prefer the simpler one.
“Real fictions” (Vaihinger): Some scientific models are false or even contradictory yet still count as knowledge. The double-slit experiment: light behaves as both a wave and a particle — a contradiction that is nonetheless scientifically productive.
Experiment and Observation
- There is no single scientific method — science is a toolkit, not a procedure.
- Serendipity plays a role: Percy Spencer (microwave oven, 1945); Penzias & Wilson (cosmic microwave background, 1964).
- Correlation ≠ causation: observation finds the pattern; theory finds the mechanism.
- Replicability: a result becomes knowledge when it can be reproduced by other researchers.
- Science is collaborative — the CRISPR discovery combined insights from microbiology (Charpentier) and structural biology (Doudna).
Theories, Laws, Models, and Assumptions
- Theory (scientific sense): a well-supported explanatory framework — not a guess.
- Law: a regularity in nature. Debate: are laws built into the universe (realist view) or useful summaries of observations (anti-realist view)?
- Model: a simplified representation. The Bohr atom models the atom as a miniature solar system — false in important ways, useful in others. Like a metaphor: it works in some respects, not all.
- Background assumptions: science cannot start without assuming regularity, causality, and the reliability of observation. These are not scientific results — they are the preconditions for science.
Theory Choice and Paradigm Shifts (Kuhn)
Five criteria for choosing between competing theories:
| Criterion | What it asks |
|---|---|
| Predictive power | Does it make accurate predictions? |
| Explanatory power | Does it explain more phenomena? |
| Consistency | Is it coherent internally and with other accepted theories? |
| Simplicity | Is it more parsimonious? |
| Fruitfulness | Does it generate new research directions? |
Thomas Kuhn — paradigm shifts: Science normally proceeds by solving puzzles within an accepted framework (normal science). When anomalies accumulate, a revolutionary shift replaces the entire framework. Kuhn compared this to a Gestalt switch — like seeing the duck-rabbit flip.
Incommensurability: Kuhn argued that paradigms cannot be directly compared — they change not just theories but the standards used to evaluate theories. This is controversial; Galison argues scientists share a “trading zone” that allows communication across paradigm shifts.
Case study — Big Bang vs. Steady State:
| Steady State | Big Bang | |
|---|---|---|
| Key prediction | Continuous creation of matter | Cosmic Microwave Background radiation |
| Falsified by | CMB discovery (Penzias & Wilson, 1964) | n/a |
Perspectives
Are There Perspectives in Science?
Yes — in almost every discipline:
- Physics: competing interpretations of dark matter, quantum gravity, string theory
- Biology: all biologists accept evolution, but disagree on mechanism — Modern Synthesis (natural selection on random mutation) vs. modern extensions (epigenetic factors)
- Chemistry: even at elementary levels, what is “really happening” in a redox reaction is interpreted differently
Science is not perspective-free. The question is whether that is a problem.
Unity of Science vs. Pluralism
| View | Core claim | Metaphor |
|---|---|---|
| Unity of science | One world → one true unified theory | A notebook: knowledge is a set of true sentences |
| Pluralism | Many maps of one territory | A patchwork quilt: different theories serve different purposes |
Pluralism is more honest about how current science actually works — theories in different domains often do not fit together well.
Falsificationism and Its Problems
Karl Popper (1902–1994): - Scientists should falsify hypotheses, not confirm them - The null hypothesis (H₀) is what gets tested; only by showing H₀ is false can H₁ be accepted - Use “support” not “proof” in scientific contexts — proof belongs in mathematics
Three serious problems with Popper:
- Self-refutation — the falsifiability principle is itself not falsifiable
- Duhem-Quine thesis — hypotheses come in bundles; an auxiliary hypothesis (e.g., “the equipment was broken”) can always absorb a falsification
- Stubborn theorists — continental drift: overwhelming evidence by 1905, accepted only in 1960. Anomalies don’t overthrow theories — they drive revision
Case study — the Periodic Table: When noble gases were discovered in the 1890s, they didn’t fit Mendeleev’s table. Scientists added a new column rather than abandoning the table. Moseley (1913) resolved remaining anomalies by reordering by atomic number.
Socio-Cultural Perspectives
Science is a social activity — who participates shapes what questions get asked.
Historical exclusion of women: - Denied education, institutional membership, and publication access - Stereotype threat: the “girls aren’t as good at science” narrative was self-reinforcing
Key figures who overcame exclusion:
| Person | Contribution |
|---|---|
| Ada Lovelace (1815–1852) | First computer programmer; designed algorithm for Babbage’s Analytical Engine |
| Marie Skłodowska-Curie (1867–1934) | Two Nobel Prizes (Physics 1903, Chemistry 1911); initially omitted from her own nomination |
| Mae C. Jemison (b. 1956) | First African-American woman in space; chemical engineering + medicine |
| Katie Bouman (b. 1989) | Algorithm produced the first image of a supermassive black hole (M87, 2019) |
Underrepresentation is not just an ethical problem — it is an epistemological one. The agenda of science (what gets studied, what gets funded) is shaped by who is in the room.
Challenges
Challenges of Observation
1. Practical equipment challenges: skill required to use instruments; reading measurements accurately; equipment malfunction.
OPERA experiment (2011): neutrinos appeared to travel faster than light — caused by a faulty fibre-optic cable.
2. Selectivity of observation: we always observe something specific, guided by theory. A circular situation: to know which variables to measure, you need to know which are relevant — which requires knowing the answer.
3. Theory-laden observation (“seeing-as”):
“There is no such thing as immaculate perception.” — Nietzsche
All observations are structured by prior concepts. A cat can see light from a laptop but cannot see it as a computer — it lacks the concept. Once you learn the constellation of Orion, you cannot see those stars as a random pattern.
Case study — Percival Lowell and Martian canals: A mistranslation (canali → “canals”) primed Lowell to observe an elaborate canal network on Mars, which he published as a 400-page book in 1906. Telescopes and space travel eventually revealed the “canals” were strings of craters. One mistranslation → decades of distorted observation by trained astronomers.
4. Observer effect and probe effect: - Observer effect: observing a system requires bouncing photons off it, changing the system’s state - Probe effect: the measuring instrument is in contact with the system (e.g., a thermometer changes the temperature of the liquid it measures)
Challenges of Testing
Confirmation bias: tendency to notice and emphasise evidence that confirms a hypothesis.
Case study — N-rays (Blondlot, 1903): Claimed to discover a new radiation. Professor Wood (sent by Nature to investigate) secretly removed an aluminium prism central to the experiment — Blondlot continued to report seeing effects. The “observations” were products of expectation, not reality.
Background assumptions: the starting assumptions of an investigation can block the correct answer.
Case study — H. pylori: Stomach ulcers were assumed to be caused by chemical imbalance because stomachs were assumed too acidic for bacteria. Marshall and Warren’s 1982 discovery of the bacterium H. pylori as the real cause was resisted for years because it violated this background assumption.
The Problem of Induction
Induction — drawing general conclusions from particular cases — is the fundamental reasoning tool of science.
Two problems: 1. No set standard for how many observations are “enough” 2. We generalise to cases we have not observed — but science is supposed to be empirical
No matter how many confirming observations you make, a counterexample is always possible.
This is why Popper proposed falsification. But falsification has its own problems (see above). Induction is inescapable.
Unobservables
Many central scientific entities cannot be observed, even in principle:
- Gravitational, electric, and magnetic fields
- Energy (only its effects can be observed)
- Quarks — constituent parts of protons; can never be isolated
Empiricists / anti-realists: only observable items exist; quarks and fields are useful fictions.
Realists: it would be a miracle if quark theory made correct predictions but quarks didn’t exist (the no-miracles argument).
This debate has not been resolved. It raises the question: is science strictly empirical in the sense of only accepting the existence of observable things — or only loosely empirical?
Key Vocabulary
| Term | Definition |
|---|---|
| Brute fact | A fact independent of any observer’s values or culture |
| Falsificationism | Popper’s criterion: a claim is scientific only if it can in principle be shown false |
| Pseudoscience | Claims that appear scientific but whose statements cannot be properly tested |
| Closed explanation | An explanation that rules out the possibility of being wrong |
| Pure science | Knowledge-making motivated by curiosity, without practical application |
| Applied science | Use of pure science to solve practical problems |
| Direction of fit | Pure science: theory fits world; applied science: world fits specification |
| Paradigm shift | Kuhn: a revolutionary replacement of one scientific framework by another |
| Normal science | Kuhn: puzzle-solving within an accepted paradigm |
| Incommensurability | Kuhn: paradigms cannot be directly compared — they change the standards of evaluation |
| Auxiliary hypothesis | An assumption that can absorb a failed experiment instead of the main hypothesis (Duhem-Quine) |
| Selectivity of observation | Choosing what to observe before the investigation — never neutral |
| Theory-laden observation | All observations are structured by prior concepts and expectations |
| Seeing-as | Perception is active: we see things as something, not neutrally |
| Observer effect | The act of observing changes the system being observed |
| Probe effect | Contact of measuring device with system changes it |
| Confirmation bias | Tendency to favour evidence that confirms a hypothesis |
| Background assumption | An assumption needed to start an investigation; can skew the inquiry |
| Problem of induction | No finite observations can guarantee a universal claim |
| Unobservables | Theoretical entities that cannot be observed even in principle (quarks, fields) |
| Unity of science | One world → one true unified theory all sciences converge on |
| Pluralism | Multiple theories can co-exist; a patchwork quilt of knowledge |
| Wave-particle duality | Light behaves as both wave and particle depending on the experiment |
| Null hypothesis (H₀) | Hypothesis of no relationship; what scientists attempt to falsify |
| Replication | Repeating an experiment to verify results; central to scientific credibility |
| Serendipity | Accidental discovery in science — valuable but not random |
The natural sciences are often contrasted with the human sciences in essays. The key comparative points: natural sciences study the world independent of human beings; human sciences study phenomena that only exist because of human agreement. Natural sciences can design controlled experiments; human sciences often cannot. The observer effect is a challenge in both — but it is central to the human sciences in a way it is not in the natural sciences.