Philosophy of Science Exam Prep PDF

1a: Induction; Logical Empiricism/ Positivism. Staley Ch. 1, Staley Ch. 4.1, 4.2 1b: Falsification, Popper. Staley Ch. 2 Popper 2a: Underdetermination; paradigms, Kuhn Staley Ch. 3 & 5 Kuhn 2b: Social constructivism; Scientific Research Programs, Lakat...

1a: Induction; Logical Empiricism/ Positivism. Staley Ch. 1, Staley Ch. 4.1, 4.2 1b: Falsification, Popper. Staley Ch. 2 Popper 2a: Underdetermination; paradigms, Kuhn Staley Ch. 3 & 5 Kuhn 2b: Social constructivism; Scientific Research Programs, Lakatos Staley Ch. 6 Mellon 3b: Realism and anti-realism; Models in Science Staley Ch.10 Potochnik et al 4a: Explanation Staley Ch. 11 4b: Feyerabend Staley Ch. 7 5a: Values Staley Ch. 12 5b: Why trust science? Oreskes 6a: Philosophy of social science: Interpretation Geertz Rosenberg 6b: Philosophy of social science: Rules, Conventions and Norms Game theory Chat prompt: Consider 'Popper Conjectures and Refutations' and Chapter 2 from Kent Staley. Could you please provide a summary of these texts, including all key points and supporting details? The summary should be comprehensive and accurately reflect the main message and arguments presented in the original text, while also being concise and easy to understand. Could you please provide a summary of Chapter 11 (excluding 11.3) from Staley, diving deep into all key points and supporting details? The summary should be comprehensive and accurately reflect the main message, arguments and concepts presented in the original text, while also being concise. Assume that the reader of the summary already has a basic understanding of philosophy of science and that he is familiar with philosophical terms; don't be superficial. Session 1a: Induction; Logical Empiricism/Positivism. Staley Ch. 1, Staley Ch. 4.1, 4.2 Deductive reasoning: if the premises are true, the conclusion is too; logical validity (which is not the same as truth, e.g. when the premises aren’t actually true). The problem of induction: using observation alone to justify general conclusions is problematic. Logical positivism/empiricism: insistence that all views must be verifiable through experiment or observation, and that all arguments must have a clear logical structure. The language of logic is much clearer than our language of speech, which has many unclarities and ambiguities. Empirical, verifiability is important. Inductive reasoning is logically invalid: a universal statement cannot strictly follow from a finite set of observations - David Hume. Answer empiricists: No certainty but probability Confirmation instead of verification. Degree of confirmation. Principle of the Uniformity of Nature: philosophical principle that asserts that the laws and regularities observed in nature will continue to hold in the future. E.g. ‘ravens 1,2,3,4 … are black’, add PUN premise; ‘All ravens are alike’. Now the argument is deductively valid, but this premise cannot be empirically justified; it isn’t a priori knowledge and also cannot be learned from experience. If-then logical operation. When you’re sick and make a promise: p; if I am better, q; I will go to class. Only break promise when you are better but do not go to class. For all other combinations, the if-then statement returns true, as you did not break your promise. Session 1b: Falsification, Popper. Staley Ch. 2 Popper Popper’s 3 motives: 1. Looking for confirmation or verification is wrong (critical of inductive reasoning) 2. Theory ladenness of observation (observations already have in them preconceived ideas of a theory) a. Pure observation does not exist; always informed by prior expectations 3. Based on deduction, logically valid Scientific theories are conjectures, not verified truths. Science progresses through a process of conjecture and refutation. A theory is scientific only if it can be potentially falsified by empirical observation. The more falsifiable a theory is - the more potential observations could prove it wrong - the more scientific it is. Induction is not a reliable theory for acquiring knowledge; it cannot guarantee the truth of its conclusions. Corroboration (evidence that supports the theory/statement) is a measure of resilience of attempted falsifications and a more appropriate way to assess the strength of a theory than verification. Formulate the theory precisely, then test it in a way that it could be refuted. Falsifiability only applies to descriptive statements; not to definitions and normative claims. Darwin: species survive that are the fittest. Not falsifiable. What is survival? What is fitness? ‘There are fish in the sea’ is not falsifiable; it needs specification. At what time and place? Problem with falsifying theories that Staley mentions: If you have a theory and set out to falsify it, the person who came up with the theory might reject counterexamples. Example of claim by person A: ‘all googly eyes are funny’. Some person B could present A with googly eyes that are not funny and person A might say ‘these are not funny, so they cannot be googly eyes, as they must be funny according to my theory.’ These things have to be agreed upon in advance. Criticism on Popper: 1. Popper says: Repeatedly passing falsifying tests (corroboration) means it is a better theory. Is this not like induction? No: with corroboration you are actively trying to falsify a theory, but failing to do so, which makes passing many of these tests a strong sign that the theory is reliable and robust to many (tested) errors. With induction, you are trying to confirm the theory, which will often lead you to find the results you are looking for, as you have a bias and might interpret results differently. 2. Falsification -> rejection. Is that not too harsh? We would throw away more theories than we would like, especially with immature or starting theories 3. Problem of underdetermination (Duhem-Quine problem) 2 & 3 are starting points for Kuhn. Session 2a: Underdetermination; Paradigms, Kuhn Staley Ch. 3 & 5 Underdetermination is the idea that evidence may not be sufficient to determine which of multiple competing theories is correct. Duhem: if we want to test a theory, what we really do is test a whole network, as you also test all kinds of background assumptions, you don’t just test a single hypothesis. Crucial experiments to test individual hypotheses in isolation are impossible because hypotheses can only be tested in combination with conditions. Scientists have to use their ‘good sense’. Quine: build on Duhem. The whole of our knowledge is a web of beliefs (critique on reductionism). Quine critiques the ‘dogma’ of the analytic-synthetic distinction. Analytic statement: true/false in virtue of the meaning of their terms (‘all bachelors are unmarried’, mathematical truths). Synthetic statements: true/false in virtue of how the world is (‘There are 8 planets in our solar system’, ‘vegetarians are less aggressive than meat-eaters’) Thomas Kuhn. Scientific progress is not a linear accumulation of knowledge. Kuhn introduces paradigms; frameworks for practising science. Elements of paradigms: theoretical principles, tools & techniques, exemplars. Theoretical principles and tools & techniques combined are also called the ‘disciplinary matrix’ (e.g. Newton: gravity, force, mathematical laws of motion etc.). Exemplar: concrete problem-solutions that students encounter in their education (e.g. Newton’s explanation of planetary motion using his law of universal gravitation). Kuhn argued that paradigms are incommensurable; no neutral standard for comparing them. There is no neutral observation language. And no straightforward translation of terms between P1 and P2. And different paradigms categorize the world differently (this explains 1&2). During periods of "normal science," scientists work within a shared paradigm, solving puzzles and refining existing theories. Exemplars serve as models for addressing new problems, and the disciplinary matrix provides the tools and concepts for doing so. However, Kuhn also posited that science undergoes periodic "revolutions" where existing paradigms are challenged and potentially replaced. These often arise when anomalies accumulate. Session 2b: Social constructivism, Scientific Research Programs, Lakatos Staley Ch. 6 Ron Mallon: ‘A Field Guide to Social Construction’ Mallon proposes that the core concept of constructionism lies in recognizing the substantial, often overlooked, influence of human choices and cultural contexts on various phenomena. Scientific theories are highly influenced by the people constructing them and authoritative persons leading the research; social construction of theories. Charisma, prestige and authority play an important role. Social dependence: Innateness and social dependence are not necessarily mutually exclusive. Something can cause you to believe; your teacher tells you about a theory, that is not a reason to believe the theory is true, it has only led you to the theory, justifying it is something else. Fact = what community of scientists agree on - social construct. Production of knowledge is forgotten/deleted; only the ‘facts’ remain. Objection1: this is implausible radical anti-realism. Objection2: What about claims of sociologists? Are these also mere constructions? Nevertheless, scientists construct theories, which do not always simply mirror nature. Ian Hacking: Objects are unaware of classification, but humans react to it. Looping effect: new knowledge becomes known to objects of classification -> changes their behavior -> changes knowledge & classification. Lakatos (back to rationality). Challenged the logical empiricist view of science as a straightforward process of accumulating knowledge. Emphasis on the historical development of scientific theories and role of theoretical commitments in shaping research programs. Sought to create a more nuanced model than Popper and Kuhn. Research Programs (SRP): Can be seen as a refinement of Kuhn’s idea of paradigms. Scientific progress occurs through competing SRPs, which are characterized by a Hard core of fundamental principles (/ irrefutable statements) that are considered unfalsifiable within the program Protective belt of modifiable auxiliary hypotheses that are modified and adjusted to accommodate anomalies and new evidence. Negative heuristic (heuristic = rule): forbids criticism of hard core Positive heuristic: suggests how to modify refutable parts in protective belt; drives the programme. SRP performs well / is progressive if it is theoretically progressive (predicts new facts and/or solves old anomalies) and empirically progressive (part of this is corroborated). Otherwise, it is degenerating / stagnating. Scientific revolution: degenerating SRP is replaced by a progressive one. When to abandon a degenerating SRP? Session 3a: Epistemological anarchism, Feyerabend Staley Ch. 7 Feyerabend: any prescribed set of rules for scientific methodology will inevitably encounter situations where violating those rules proves more conducive to the advancement of scientific knowledge. Epistemological anarchism: there is no singular, universally applicable scientific method; he advocates for a flexible and adaptable approach to science; ‘anything goes’. Science is a creative process, like art. Applying rules requires context. Pluralism. Use counterinduction. Scientific knowledge shouldn’t be viewed as a linear progression toward a single, unified truth, but rather as an ‘ocean’ of diverse and potentially conflicting perspectives, which can co-exist. Humanitarian attitude towards science emphasizes individual freedom (influenced by Mill). Problem: his anarchistic approach could lead to issues like the influence of non-scientific factors in research. Session 4a: Explanation Staley Ch. 11 Hempel & Oppenheim’s covering-law (CL) model: an explanation consists in show that the phenomenon to be explained (explanandum) can be logically deduced from a set of general laws and relevant initial conditions (explanans). Two variants: Deductive-Nomological (D-N) explanation (Hempel & Oppenheim): universal laws. Explanandum is logically derived from a set of premises; explanans. Explanations take the form of deductive arguments. Inductive-Statistical (I-S) explanation: probabilistic laws. The explanans does not deductively entail the explanandum, but instead establishes a high probability for its occurrence Causal theories of explanation. Explanation = uncovering the causes of phenomena. What is causality? Four answers: Regularity theories: based on the idea that causality is a product of our mind (Hume Treatise, positivists) ○ Avoid speculation about inherent ‘powers’, ‘dispositions’ ○ Logical positivism: causes are metaphysical, delete this talk. OR when we talk of causality, we really mean to indicate regularities. ○ Regularity view of causality: covering law account of explanation Counterfactual theories: based on the intuition that if the cause hadn't occurred, the effect wouldn’t have occurred either (Hume enquiries, Lewis, Woodward) ○ Model Woodward: X is a cause of Y if: Temporal succession: X is prior to Y Robust correlation: X and Y correlate Counterfactuality: no X => no Y Process theories: based on the idea that one can identify causal processes and interaction in nature Mechanism theories: (not important for exam) Session 4b: Realism and anti-realism; Models in science Staley Ch. 10 Potochnik Realism: Scientific theory gives a true description of reality behind our observations. The goal of science is to explain phenomena by means of a true theory. All claims and concept relate to something in the world Instrumentalism: Scientific theory is merely an instrument to summarize observable phenomena. The goal of science is to describe and predict phenomena by means of an empirically adequate theory. Logical positivists were instrumentalists: Verification principle: claims about reality which are speculative and not verifiable should be banned. Distinction between observable terms and theoretical terms. Arguments pro realism: No miracles; inference to the best explanation. The success of science, particularly its ability to predict novel phenomena, would be highly improbable or even miraculous if our scientific theories were not at least approximately true. Distinction observable-non-observable is not sharp: This argument, while not directly supporting the truth of scientific theories, challenges a key premise of anti-realism. Anti-realists like van Fraassen argue that we should only believe in what is observable. However, proponents of realism point out that the distinction between observable and unobservable is not always clear. Reaction: the extremes are very clear, only the border between them might be vague, but this does not mean that we cannot generally distinguish these things. Pro instrumentalism: Underdetermination: The available evidence often underdetermines our choice between different theories. Multiple theories might make the same predictions about observable phenomena but differ in their claims about unobservable things. Therefore, instrumentalists argue, we shouldn't believe in the truth of any particular theory, but only in its usefulness as a tool for prediction and explanation. Pessimistic meta induction: points to the history of science, where many successful theories, once thought to be true, were later replaced by radically different theories. The pessimistic meta-induction concludes that, because past successful theories have been proven false, we have no reason to believe that our current successful theories are true. Models in science (Potochnik) Models serve three primary functions: Representing: models stand in for a target system, encompassing a range of range of relationships between model and target, enabling us to study and learn about it. Directing: Models guide research by focussing on specific aspects of the target system and suggesting experimental interventions Describing: Models provide a framework for describing the target system and for communicating scientific findings. Models can describe even when they don't directly represent. … Session 5a: Values in science Staley Ch. 12 Epistemic values: contribute to the pursuit of knowledge and truth; accuracy, simplicity, explanatory power, consistency, scope, fruitfulness (Kuhn, paradigm-independent rational) Non-epistemic values: moral, social, political and economic considerations, raising questions about their proper role in science. The inductive risk argument: non-epistemic values can legitimately influence scientific decision-making. This carries the risk of making incorrect inferences: Accepting a false hypothesis (false positive, type I error), or rejecting a true hypothesis (false negative, type II error). If this forms a large risk, these values can inform scientists so they might set a higher threshold of evidence for accepting that hypothesis. Merton’s norms of science, for effective functioning of the scientific community: Universalism: scientific claims should be evaluated based on their merit, regardless of the researcher’s personal attributes or social standing Communism: Scientific knowledge should be shared openly and freely within the community Disinterestedness: Scientists should not be motivated by personal gain or ideology, but by the pursuit of knowledge. Organized skepticism: all scientific claims should be subjected to critical examination Feminist critique: ‘value-free science’ is influenced by gender bias. These values/biases should be addressed and acknowledged, not pretended to not exist. Plagiarism. People manipulate or fabricate data to match their research. Experiment/test results are omitted if they don’t match the research hypotheses. How to establish whether someone has co-authored a paper and the order of authors? Conflicts of interest. Funding bias Remedies for lack of objectivity: Pluralistic objectivity: alternative to objectivity as value-neutrality: objectivity is better achieved through a critical and diverse scientific community that encourages the open exchange of ideas and perspectives Session 5b: Why trust science? Naomi Oreskes: ‘Why Trust Science?’ Many instances where scientific claims, initially accepted as facts, were later overturned. This raises questions about the reliability of scientific knowledge and how we evaluate the truth claims of science. Oreskes discusses the following factors that make science trustworthy: Methodological standards, peer review, transparency, consensus, and responsiveness to evidence. People speak of ‘post-truth’ (the idea that objective facts are less influential in shaping public opinion than appeals to emotion and personal belief), ‘scientists construct ideas and theories’, ‘and they disagree all the time’. Kuhn and sociologists: theory choice is a matter of socio psychological factors. Why trust science? Five key themes for ensuring reliable knowledge production Expert consensus: It is not true that scientists disagree on everything. Underneath their differences is a broad consensus on many things. Disagreeing non-experts are, as a rule, less relevant. Method: Don’t be too obsessed with just one method. Embrace pluralism. Emphasizing the importance of rigorous methods and experimental design, while being aware of limitations and biases. Evidence: comprehensive and balanced evaluation of evidence, acknowledging uncertainties. Values: look for confirmation while avoiding confirmation bias, do not ignore counter evidence, be open to falsification. Recognize the influence of epistemic and non-epistemic values. Humility: acknowledging the possibility of error, remaining open to challenges and revisions of existing knowledge Diverse teams: multiple perspectives, demographic diversity serves as a proxy Session 6a: Philosophy of social science; interpretation Geertz: Thick Description Rosenberg: The Explanation of Human Action Geertz: culture is best understood as a system of meaningful symbols (webs of significance) that people use to understand and interact with their world; interpretive approach to anthropology. Thick description: not just describing what people do, but interpreting those actions in light of the cultural contexts in which they occur (wink, parody of a wink, or twitch). He criticizes approaches to anthropology that rely on thin description, which simply records behaviors without considering their cultural significance. Rosenberg: focusses on the role of interpretation in explaining human action, arguing that we must understand the beliefs and desires that motivate people in order to explain what they do. Distinguishes between ‘mere behavior’(e.g. the beating of our hearts) and action (behavior that is under our control). Two main mental state categories: Beliefs: mind-to-world direction of fit; mental content must mirror the world. Desires: world-to-mind direction of fit; world must mirror mental content. I do not have any beer, I order a beer and it comes to me; the world changes, comes to match my mental state. Desire is like a shopping list, belief is like an inventory list. Basic assumption: actors are rational and act on mental states. These beliefs and desires are internally consistent, make up a coherent set. Utility scales and axioms: Ordinal utility: only concerned with the order/rank of utility of possibilities ( b-z-g medals), nothing about quantity. Preference must satisfy 3 axioms: ○ Asymmetry: cannot prefer wine over beer and beer over wine at the same time ○ Completeness: for all pairs of possibilities A and B, an agent either prefers A over B, B over A, or is indifferent. People make non-random choices ○ Transitivity: A > B, B > C => A > C. Interval utility function: two more axioms needed if we also want to know how close alternatives are to each other (example interval scale: temperature). Interval scale allows positive linear transformations (of form ax + b). Intervals are the same, ratios are not the same (Celcius and Fahrenheit) ○ Independence: addition of an alternative does not change relative relations of others (waitress says: chocolate pie or apple pie. I prefer chocolate pie. Waitress returns, forgot to mention carrot pie, now I suddenly prefer apple pie. This does not make sense. ○ Continuity. I prefer €1,000,000 over €0 over death. There exist a unique probability p such that p(€1,000,000) + (1-p)(death) = 1(0). Indifferent between having a (large) probability of getting a million euros with (small) probability of dying, and getting 0 euros with certainty. Ratio scales: intervals are the same and ratios are the same (centimeters and inches), allows transformation in the form ax, the zero is fixed. We often only have behavioral data, without straightforward data about underlying preferences or motives/desires of subjects. Stated preferences are not always reliable; underdetermination problem; behavior data underdetermines hypothesis. Try to estimate these thoughts, postponing the conclusion that the person is not rational for as long as possible; try to explain their behavior. Session 6b: Philosophy of social science; games, equilibria and norms Gilboa: Game Theory Nash Equilibrium: no player’s outcome can be improved by changing his own strategy. A combination of strategies is an NE if neither party has a reason to unilaterally change his strategy Pareto Efficiency: impossibility of improving at least one player’s outcome without harming any other outcomes. Dominant strategy: a superior strategy regardless of how the other player(s) act(s). In prisoner's dilemma; always confess (thumbs down). When both players have a strictly dominant strategy, where the combination of both results in a NE, we have a dominant strategy equilibrium. Stag hunt example. Hunt stag or hare, need two people to hunt stag, only one to hunt hare. nth order beliefs. For A, S is a good action if B wills S (1st order belief), A must also believe that B believes that A wills S (2nd order belief). How to ensure both parties choose optimal option for both? Convention, rule, norm: behavioral regularity, deviant behavior gets criticized. Empiricism is a theory of knowledge that asserts that knowledge comes primarily from sensory experience. It emphasizes the role of empirical evidence in the formation of ideas, rather than innate ideas or traditions Naturalism, on the other hand, emphasizes the unity of the natural world and the methods of science. It asserts that all phenomena can be explained by natural laws and processes, without the need for supernatural explanations. Naturalists believe that science should focus exclusively on the natural world and should not invoke supernatural entities or forces in its explanations. Realism: Scientific theory gives a true description of reality behind our observations. The goal of science is to explain phenomena by means of a true theory Instrumentalism: Scientific theory is merely an instrument to summarize observable phenomena. The goal of science is to describe and predict phenomena by means of an empirically adequate theory. (Constructive Empiricism) Logical positivism is a philosophical approach that emphasizes the importance of empirical evidence in determining the truth of propositions. It rejects metaphysical and idealistic claims, and asserts that knowledge can only be derived from scientifically confirmed statements. It considers that the only meaningful philosophical problems are those which can be solved by logical analysis.