Phil360 – Astronomy – Flashcards

• What does Feynman mean by "Cargo cult science" and by "integrity in science"? Do you agree?Is there something missing from Feynman's account of science? o Cargo Cult science Refers to the practices of tribes who thoughts that if they replicated conditions under which planes came bearing materials during the war, then the planes would come again with such materials. These conditions included runways, fires at the side, wooden for a man to sit in, and bars of bamboo sticking out of the head like antennas The ideas is that although the form under which the planes would land was present, they were "missing something essential" - the planes. Feynman applies this principle to endeavours which claim to be scientific and bear the form of science, but are "missing something essential" - "one feature" is scientific integrity. o Scientific Integrity It's a principle that corresponds to a kind of "utter honesty-a leaning over backwards" If, for example, you are doing an experiment, then you should report everything that might make it invalid, including • alternative causal explanations for the results, • prior explanations which you think have been eliminated and • data which may throw doubt on your preferred interpretation. Refer to previous experiments Don't fool yourself or others [scientists, laypersons etc.], Publish all results "try to give all of the information to help others judge the value of your contribution; not just the information that leads to judgment in one particular direction or another" • Do you agree? • Is there something missing from Feynman's account of science? o Yes there is, but he acknowledges it Scientific enquiry is not just about you sharing all the knowledge and practices you have had to other scientists, but making sure you have all the shared knowledge and practices which scientific inquiry takes place in However, Feynman says that there is "one feature that I notice that is generally missing in Cargo Cult Science," not "only one" feature missing.

1. P is conclusively verified = P has actually been shown to be true (or P follows from a finite set of consistent set of true observation reports that have actually been made). a. If P has been verified, then -P has been falsified. 2. P is conclusively falsified = P has actually been shown to be false (the falsity of P logically follows from true observations). 3. P has been conclusively tested = P has been verified or falsified conclusively. 4. P is conclusively verifiable = S could be shown to be true or verified (or a humans could make a finite set of observations which would entail the truth of P). 5. P is conclusively falsifiable = P could be shown to be false. 6. P is conclusively testable = P is conclusively verifiable or falsifiable.

1. Logically possible = S is permitted by the laws of logic and rules of language 2. Naturally or physically possible = S is permitted by the laws of nature 3. Technically possible = S is permitted by our current human abilities to do things

Testing Singular Observational Statements 1. Singular statements: ones which refer to one object 2. Reports of current observations are conclusively verifiable and falsifiable 3. Reports about past observables: a. One can infer to past events on the basis of two things: i. 1) laws and generalisations, and ii. 2) observations about the present (archaeological traces and testimonies) b. Here is an argument type: i. Current observation: O is a report of some observable(s) ii. Law or generalisation: P generally causes the phenomenon of O iii. If O and P generally causes O, then P existed iv. Therefore, P existed c. Example: i. There is some ash ii. Fires generally cause ash iii. If there is some ash and fires generally cause ash, then a fire existed iv. Therefore, a fire existed d. Evolution and dinosaurs are included in scientific theories.

1. These begin with universal statements "all" "every" 2. Testability: a. They can be verified by showing all such objects meet it b. They can be falisified via a counter-example 3. Universal or general statements can be restricted or unrestricted. a. Restricted claim: in that it refers to a finite space, time and number - number of objects in a finite amount of space-time b. Unrestricted claims: If it was infinite, one could not complete that search. It would not be conclusively verified, but could be conclusively verified. i. Laws of nature are not verifiable, but they are falsifiable since they involve unrestricted claims 4. Confirmation means large number of examples in its favour and no counter-examples - admits of degrees

1. Examples a. There exists at least one person in the room under three metres tall 2. Distinction a. Restricted existential claims: Refer to a finite portion of space-time i. There is a lockness type monster in this lake ii. Jesus will be resurrected in two days in France b. Unrestricted existential claims: Refer to an infinite portion of space time i. There is a lockness monster somewhere ii. Jesus will be resurrected somewhere at some time 3. Testability a. Restricted claims are verifiable or falsifiable b. Unrestricted claims are verifiable but not falsifiable 4. Translatability a. Unrestricted existential claims can be translated into negated unrestricted universal claims

Testing mixed quantified statements: • Example: o There exists a person who everyone loves o There is a substance which dissolves every other substance o Causal principle: For every event E (not verifiable), there is another event C (not falsifiable), such that C causes E o Principle of Universal explainability or Sufficient Reason: For every E event E, there is some law or theory T such that T explains E o Principle of Scientific Materialism: For every mental process M, there is some physical process P such that P explains M o Principle of Creationism: Everything is created from some other thing o Newton's Law: For every action A1, there is another action A2 such that A1 is equal and opposite to A2 and A1 is a reaction to A2. Pressing on a desk (debatable as the existential claim could be restricted) o Principle of Conservation of Energy: Any increase in energy in an isolated system corresponds to a decrease in energy or momentum in another part o Principles of Newton and Conservation are presupposed and help physicists o Biological adaptationism: For every organ, feature or behaviour of individuals of any species, there is some function or purpose which it fulfils and for which evolved. (Adaptionists shoud allow that organs can lose their funtion) • Cannot be verified or falsified = They seem to be beyond testability.

Relatively testable and absolutely untestable statements: • Absolutely untestable hypotheses U may be relatively testable if there is some testable statement or theory T whose truth supports or is inconsistent with T>(-UvU) • The causal principle can be relatively testable since quantum mechanics is testable, and it could imply the falsity of CP o QM is confirmed>-Cp o QM is confirmed o Therefore, -CP • Some claims are untestable o Five minute creationism o Size Doubling

Testability as an aim of science is problematic • If science seeks truth about untestable propositions, then this is problematic E.g. gravity

1. Science aspire to more than just reports of what we currently observe 2. We make inferences about unobservables a. The past i. Making claims about the past is also called retrodiction in contrast to prediction b. The future i. Prediction c. Elsewhere d. Laws of nature 3. Newton gave it the name induction

• Do you think there are claims in science which are not testable (in a sense you need to define)? If so, are they to be rejected from science, or do we have to alter our conceptions of what counts as testable? If so, what would that be? o Principle of Universal explainability or Sufficient Reason: For every E event E, there is some law or theory T such that T explains E o Quantum level exclusive o Why it is untestable o Implicit claims • If so, are they to be rejected from science, or do we have to alter our conceptions of what counts as testable? If so, what would that be? o This is an essential claim in science. Scientific enquiry is the search of explanations. If such explanations are not presumed to exist, then scientific inquiry halts o Therefore, we need to alter our conceptions of science as a pursuit which may not only assume testable assumptions, but can involve certain untestable assumptions. o Which assumptions those are are beyond the scope of this work.

• Claim 1. We are to admit into science only the following Reports of observations Any statement (deductively or inductively) inferred from these observational reports 2. We are to reject all else as non-science, pseudo-science, mere speculation, etc. 3. In response to theory-laden observational objection: All observation is to be independent of any theory and must not be theory-laden, or presuppose any theory. • Weaker Naïve inductivism could eliminate theory-laden condition since induction from theory laden observations can still work.

1. Demarcation of science from non-science: Unscientific beliefs are ones arrived at without proper observations and inductive inferences. 2. Method of discovering laws and theories: 3. Theory of conditions for justifications: We are justified in accepting a scientific theory if we can show that it has been arrived at via inductive inference from observation

• These are inferences in which you enumerate a number of cases and then draw a conclusion about the next or unobserved cases • Predictive Induction - infer the next unobserved case • Inductive generalisation - infer to all the unobserved cases • Inductive generalisations can be restricted or unrestricted. • All emeralds are green - Inductive generalistion • Inductive inferences only work when you have an independent way of identifying what X is without inferring Y where Y is a part of X's definition

• Analytic statement: a statement made true by virtue of logic and semantics • Otherwise it is a non-analytic or empirical statement

• Inductivists - Follow rule of induction • Counter-inductivists - Going against the rule of induction • Indifferentist - Indifferent about the conclusion from inferences • Frequential probabilities are often based on induction • Hume said life would not go on without induction

• Inductive arguments could become valid by stating the PUN • PUN could distinguish good inductive inferences from bad inferences • Hume's statement of PUN o Instances, of which we have had no experience, resemble those of which we have had experience o The course of nature continues always uniformly the same • Question regarding an instance of uniformity of nature • Nola holds that PUN is necessary to justify inductive inference but not sufficient since it does not distinguish observed regularities into those that are instances of genuine uniformities and those that are not • PUN also may or may not hold for probabilistic uniformities (e.g. it rains 25% of the time)

What does Hume say on the lack of justification of Inductive inference and scepticism? • The rationality of human argument • We humans argue in only one of two ways; Deductively and inductively (i.e., non-deductively) 1. We do argue inductively and perhaps cannot not do so. It is a deeply ingrained habit or custom. 2. There is no rational justification for induction. Attempts to justify it are circular. • Structure of Induction: • Some past observations we have made: Sun rising each day; drinking water quenching thirst, etc. • By PUN, future observations will resemble these: a future water drinking will quench thirst rather than not; etc. • Premise: We have observed that all past As are B's; • Conclusion: so, next A is a B, or all future A's are Bs. • Hume's Supposition (Enquiry #30) of Circularity • 'All our experimental conclusions [conclusions of form next A is B, or all A are B] proceed on the supposition [of PUN] that the future will be conformable to the past' [viz., all past A's are Bs] • 'To endeavour, therefore, the proof of this last supposition by probable arguments, or arguments regarding existence, must be evidently going in a circle, and taking that for granted, which is the very point in question.' • Hume's rejection of PUN as being derived from meaning or necessary truths: • it is logically possible that the future does not resemble the past, so it is not a necessary truth • i.e., it is logically possible that all observed A's are Bs but yet the next A is not a B; • Or: it is logically possible that all observed A's are Bs but it is not the case that all unobserved As are Bs. • So PUN is not a truth of meaning or necessary truth. • So it is empirical claim to be established on the basis of experience. • Hume's Argument for Circularity 1) Arguments from past observations (all observed As are Bs) proceed on the supposition of PUN (Hume's assumption about role of PUN in inductive reasoning) 2) So, we have no reason to believe the conclusion of an inductive argument (e.g., the next A is a B) unless we also believe that PUN is true. [Follows from (1)?] 3) The proposition 'Nature is uniform' (PUN) cannot be shown true by deduction as its negation is not self-contradictory. 4) Either PUN is shown by deduction or induction 5) So, PUN can only be shown by induction [(4) and (3)] 5) So, PUN (as a conclusion) can only be shown by induction 2) Unless we believe that PUN is true we have no reason to believe the conclusion of an inductive argument. 6) Unless we believe that PUN is true we have no reason to believe an inductive argument the conclusion of which is PUN (Nature is uniform] 7) If a proposition (e.g., PUN) cannot be shown deductively or inductively then we have no reason to believe it. 8) So we have reason to believe PUN [by (3) & (7)] 9) So we have no reason to believe the conclusion of any inductive argument ( since by (1) they all proceed on the supposition of PUN) (Nola doesn't think you need to believe a supposition)

What is The Circularity Problem and some responses? - Circularity problem • P1 - Can only believe the conclusions of inductive arguments to be correct on assumption of PUN. • P2 - PUN can only be shown inductively, not deductively • Therefore, Circularity: To establish PUN as a conclusion of inductive argument we need to assume it [proceed on its presupposition] • Popper agrees that there is no justification, but he disagrees that it is a habit. We may learn in non-inductive ways (without repitions), for example, puppies smoke a cigarette just once and then they dislike it. No repetition. - Rule Circularity: o Premise circularity: Assuming PUN as a premise in an argument leading to pun as a conclusion is bad o Rule circularity: Assuming PUN as a rule to establish PUN as a conclusion - not so bad. Some think it is OK. o There is rule circularity in deductive logic o Rule circularity Argument form: Premises of Inductive argument PUN = Nature is uniform - Hierarchy or levels of inductive rules: • Hierarchy or Levels of Inductive Rules 1. All observed As are Bs; 2. All unobserved As are Bs • This is a level 1 inductive rule of science. Assigns probabilities to conclusions of level 1 arguments. • Justified by a higher level 2 rule which says: I. (Most) Rules of inductive Inference at Level 1 have worked in the past II. Level 1 rules will work in the future. • This is a Level 2 rule that justifies level 1 argument. Avoid circularity. But trade this for regress of justifying rules. Harmless or problematic?

What is Naïve Inductivism, its contrast to other views, Hempel's model, objections to the model and an example of the model? • A "bottom-up" Theory of method or test where you start with observations and make inferences, theories, laws, generalisations etc • Contra hypo-thetico deductivism: starts with hypotheses • Popper's view: o Start with problem, o propose theory, o find errors in the theory o that creates a problem and then the process continues all over again o Doesn't start with observations or theory's but a problem - conflict between what we think is the case and, upon reflection, find it to not be the case • Hempel's Model of I: 1. Collection of all observational facts independently of an selection process 2. The analysis and classification of observational facts independently of any theory 3. Inductive inference from the classified facts to generalisations 4. Further testing by deductions from the generalisations • Objections to the model: o 1. It is impossible to collect all facts It is impossible to start it o 2. Theory dictates which classifications are important given the infinitude of classes • Example of the model o Investigators sought to see what caused SIDS - Sudden Infant Death Syndrome o Investigstors drew up a possible list of causes o They constructed lengthy questionnaires to identify any possible causes irrespective of plausibility o It surprised them to see that sleeping in prone position was correlated with cot death o It has now been known to reduce SIDS in other countries o There was very little filtering of facts in the process

What is the problem of justifying induction? 1. Induction is making claims about unobserved cases based on one or more observed cases. 2. It relies on the principle of the uniformity of nature a. The claim that nature is uniform in some respects b. If this is the case, then we can infer from observed uniformities to facts about unobserved cases of those uniformities 3. Argument for circularity a. PUN is either analytic or empirical b. It is not analytic c. Therefore, it is empirical d. If it is empirical, then it must be inferred inductively from observed cases. This inference presupposes that induction works. e. But if justifying induction requires that it must presuppose that induction works, then the justification for induction is circular since it presupposes what it intends to prove. This justification is a fallacy of rationality. f. Hence, there is no rational justification of induction which does not commit the fallacy. State as clearly as you can why there is a circularity problem about the justification. Can it be solved? • It can be solved by assuming foundational, basic beliefs which do not need justification • We do this all the time • Justification of induction is circular, just like the justification of memory is circular • This is a plausible solution, but it creates a problem. • What is to distinguish properly basic beliefs from non-proper basic beliefs. • This might then create the problem of justification again, but perhaps it is a start.

• Outline o Method of justification o Process o Four Definitions o Auxiliaries o Confirmation o Kinds of evidence o Consilience o Induction in the HD schema o Conjunctive confirmation or falsification o Example • HD method is a method of justifying or testing hypotheses, not discovering them • General process: o Start with a hypothesis - hunch or conjecture as Popper says - hypothetico part (and some auxiliaries) o Deduces consequences o Consequences are compared with evidence - inconsistent, independent or identical/supporting C - from observation or experimentation or from other plausible theories or laws. • Definitions o Hypothesis = any statement, whether a generalisation or a particular claim o Evidence E = conjunction of statements to test a hypothesis - reports of observations, theories or laws o Distinction between passes and fails H passes a test when E is supporting or identical with C H fails a test when E is inconsistent with C • Auxiliaries for HD o Always a hypothesis requires auxiliaries in order to be testable o IC - Initial condition of state of a system - This horseshoe is metal o M - Model idealisation or partial representation of a real system - Planets as particles etc. o B - Background theory relating to hypothesis such as law of motion o K - Addition theory in gathering evidence - thermometers relying on the theory of the expansion of theory in confined space • Confirmation o Confirms if h, e >1/2 etc. Disconfirms if h, e <1/2 etc. Neither confirms or disconfirms if h, e =1/2 etc. • Comparative degrees of confirmation and quantitative degrees • Kinds of evidence o Designer evidence - Theory T was designed with evidence E in mind o Non-Designer evidence - E was known and T accounts for E but E was not used to design T o Novel evidence - Where T implies C and C is not known as true and E confirms C. • William Whewell on Consilience and Evidential Support o Consilience occurs in supporting a theory when: Theory is supported by multiple classes of phenomena Theory successful predicts evidence of a different kind from that considered in its formulation - Non-designer evidence Successful prediction of novel evidence or things we would not have expected to occur given background knowledge - Novel evidence Theory encompasses different kinds of fact that were thought to be of quite a different kind • Induction in the HD schema o H can only get inductive confirmation from its passes o One inductively infer the replicability of the E's consistency with C • HD schema, failure and falsification, passes and inductive confirmation o Passes and failures only confirm or disconfirm the conjunction, not the single hypothesis o Duhem-quine thesis is the no H can be sufficiently isolated from A to be tested by itself o Popper thinks you could test auxiliaries and infer H must be false if others are not • Contra: Geometry goes from principles to conclusion • Probability and confirmation feature in Huygen's model, not verification as defined earlier • Nola thinks it lacks a decent theory of confirmation • Example of HD schema • (Moving Plate Hypothesis) • Data: The oldest volcanic islands are located the farthest from the hot spot and have little to no volcanic activity. In contrast, closer to the hot spot, the younger the volcanic islands are and the greater the amount of volcanic activity. For example, the island of Hawaii is presently positioned directly above the hot spot and experiences volcanic activity on a regular basis. • Explanatory Hypothesis H: The Hawaiian Island chain formed as a result of the Pacific plate moving over a fixed hot spot located in the mantle beneath the pacific ocean. I. MPH II. Speed of moving plate • Deduce/explain some known but, so far, not explained facts: 1. Order of islands in terms of age analysis of rocks). 2. Absence of volcanic activity away from youngest active volcano I. MPH II. Order of Islands • Deduce/predict the unknown: 1. Speed of moving plate • [Check this fact independently - how? If right then confirmation of MPH]

o Four: HIHO o Holist objection Can't isolate hypothesis from the auxiliaries so as to figure out what it supports or falsifies Duhem-Quine Thesis that no hypothesis can be sufficiently isolated to be tested by itself in the absence of any other auxiliaries Tycho Brahe claimed to falsify the Copernican theory that the earth revolved around the sun by observing no difference between the position of the stars at different times of the year stellar parallax The problem was with his auxiliary assumption regarding the distance of the stars o Irrelevance objection If E confirms H and A, then E confirms H, A and I where I is an irrelevant proposition HD needs an answer to why a piece of evidence does not confirm every theory that implies it. HD confirmation gives no account of evidential relevance o History objection Many true theories of history were falsified at the time but not abandoned and they consequently were proved true E.g. Copernicus • Copernicus proposed that the earth moved around the sun • Falsifications: TOMS o Tower argument - Tower argument of a stone falling to the bottom, not to the side o Objects Movement- Why don't things on the earth's surface move or o Moon movement - Why doesn't the moon fly away Gallileo: • Observed that Jupiter had moons, moved and the moons didn't go away o Stellar - No stellar Parrallax • Copernicus had no answer • The progress didn't occur through a conjecture being refuted (falsificationism) or careful observation and experiment o -(O&H) Observation or Hypothesis falsification Logic alone doesn't say whether the observation is false, or the hypothesis If falsificationist consider the observations to be fallible and theory laden, then the observation may be wrong and may not straightforwardly rule out the theory

Karl Popper: • Outline: o Two questions o Three points in conjectures o Definitions of Failsifiability o Examples of Unfalsifiable claims o Comparative degree of falsifiability o Induction o Background to Corroboration o Ideas behind probability o Likelihoods o Definition of a severity of a test o Definition of Corroboration o Corroboration and induction o Differences between Popper and HD o Conditions for falsification • Two questions as a student: o When is a theory true or acceptable? o When is a theory scientific? How do you distinguish science from other things non-science, philosophy etc. demarcation project. Newtonian mechanics is false but scientific Need definition for scientific theories which may or may not be false Thinks that theories can never be showed to be true because they are generalised But we can falsify them Karl Popper is impressed by Einstein's predictive power • Points in conjectures: o It is easy to look for confirmations for nearly every theory if we only look for them o Confirmation only counts for risky predictions which could be false o The more a theory says x will not happen, the better it is - degree of falsifiability • Falsifiability o A theory is scientific if it has at least one potential falsifier but does not have every potential falsifier (D.Fsb=1) o Theory T rules some observations out and permits others. o Theory should make a cut through all observations - consistent and inconsistent. Inconsistent obersavtions are called potential falsifiers. o If there is an actual falsifier, then the theory is falsified. o Can have many potential falsifiers and none of them are actual o A theory is testable and scientific if and only if (T&(IC&M&B)>C) • Examples of Unfalsifiable claims - Ones which have a degree of falsifiably = to O o Tautologies - Law of excluded middle is unfasifiable since it has no potential falsifiers o Definitions - are unfalsifiable o Unrestricted existential claims - Jesus will return o Abstract Mathematics - o Metaphysical claims • Comparative degrees of falsifiability 1. If A entails B, then A>B where A has a higher degree of falsifiability than B 2. If T1 has a greater degree of empirical content with more consequences than T2, then T1>T2 wrt falsifiability 3. If T1 has a lower degree of absolute probability (or prior probability) than T2, then T1 has a higher degree of falsifiability than T2 E.g. T1 = it will rain today and tomorrow, T2 = it will rain today o This is important since Popper defines scientific theories in terms of degrees of falsifiability o According to Popper, we should eschew theories which, in comparison, are not highly testable or are hardly testable at all o It is difficult to give an account of absolute falsifiability though o Falsifiability of a theory is inverse with its absolute probability • Induction and confirmation o Popper accepted Hume's rejection of the justifiability of induction but sought to find an alternative non-inductivist account of how evidence related to hypotheses. o Rejected term 'confirmation' as being too inductivist • Intro to Corroboration o In contrast provided an account of corroboration which is non-inductivist. o Corroboration can be partially defined in terms of probabilities. No full formal definition can be given. • Ideas behind probability o Probability to be defined as a two-place relation holding between evidence E and hypothesis H (or a theory T). o We commonly ask about: The probability of hypotheses H given evidence E. o Abbreviate: p(H, E). o Probabilities range between 0 and 1. 0 £ p(H, E) £ 1 o But can have comparative probabilities. If E1 and E2 are two bits of evidence we may wish to know: p(H, E1) ³ p(H, E1&E2)? Or p(H, E1) £ p(H, E1&E2)? • Likelihoods - Expectedness o In p(H, E) we can think of H as a constant, given hypothesis and let E vary as evidence comes in over time. o Reverse order of H and E: p(E, H). o We think of E as a given body of evidence and let H vary over a range of hypotheses. o We are asking how probable is fixed E for different hypotheses H1, H2, H3, etc. o This is still a probability - but of a special sort called a likelihood. [R. A Fisher & Popper] o Probability of E given H = likelihood of H on E • Severity of a test: • Popper defines the severity of a test by likelihoods of the evidence • Severity Test of H (new theory) wrt B (background theory or beliefs) by E • S(H, B,&E) = Defn p(E, H) - p(E, B) 1. H predicts E • p(E, H) ≈ 1 2. Either: B predicts E, • p(E, B) ≈ 1 • So S(H, B, E) or severity of the test is ≈ 1-1 = 0 • Minimum severity of test 3. Or: B fails to predict E, and is surprising from point of view of B • p(E, B) ≈ 0 • So S(H, B, E) ≈ 1-0 = 1. • Maximum severity of test • Difference to HD - Historical look at how predictions were prior to and with the theory • Corroboration o Corroboration is a measure of the degree to which a hypothesis/theory has withstood severe tests up until now. It is non-inductive; there is no guarantee about corroboration in the future. o Types of corroboration: Maximum positive corroboration - pass severe test Maximum negative corroboration - fail a severe test Neutral corroboration - not a severe test. o Corroboration(T, B, E) µ (is proportional to) Severity(T, B, E) o µ p(E, T) - p(E, B) o Measure of Corroboration 1: Passes Test o To get a measure of Corr. between -1 and +1 Popper uses a "normalizing" factor in the denominator: o Corroboration(H, B, E) = p(E, H) - p(E, B) o p(E, H) + p(E, B) • Case 1: Suppose E follows from H (with auxiliaries assumed), i.e., p(E, H) = 1 o Then above formula reduces to o 1 - p(E, B) o 1+ p(E, B) o Case 1(a): If p(E, B) is low or 0 (i.e., severe test of H) then Corr(H, B, E) = 1. o Case 1(b): If p(E, B) is high or 1 then Corr is 0. o So Case1(b) is a pass for H but not a severe test. • Corroboration 2: Fails test o Corroboration(H, B, E) = o p(E, H) - p(E, B) o p(E, H) + p(E, B) o Case 2: Suppose observed evidence E is inconsistent with H (with auxiliaries assumed), i.e., H is refuted by E in the severe test so that p(E, H) = 0 o Then above formula reduces to o 0 - p(E, B) o 0 + p(E, B) o = - p(E, B) o + p(E, B) o = -1 o Failing a severe test yields Corr(H, B E) = -1. • Corroboration & Induction o Corr. is report on how well a hypothesis has withstood severe tests. NOT ABOUT TRUTH SINCE THAT LEADS TO INDUCTION. Corr can be high (1), low (-1), neutral (0) o If a theory has high Corr. i.e., has withstood severe tests, can we infer that it will withstand further severe tests tomorrow? (Induction!) o Not if it is a very different kind of severe test; it might fail such new and different tests tomorrow. o But what if the severe test made yesterday is repeated tomorrow? Should we infer that it will pass or fail? o A whiff of induction suggests that we should claim that it will pass tomorrow rather than fail. o Popper's measure of corroboration is interesting, and fits some intuitions about H-D method. But we should not conclude that it helps us evade induction entirely. • Why want corroboration - Induction: o Popper says that it means it should not be rejected and it has to be accepted as perhaps empirical truth, perhaps, but it might fail in the future o We don't start with observations and then go to the theory, but rather we start with a hyporhesis from the geinning, but he calls them expectations, e.g. of regularities o We you use expectations of regularities and then use science to impose them on the world o Why do we want highly corroborated rather that not o "Because we search for truth and because the falsified theories are known or believed to be false, while the non-falsified theories may still be true..." • How do we jump to a good theory from observation? o Start with an observation o Jump to any theory and then test them • Difference between HD & Popper o Although popper uses the HD method of test, he adds some twists Falsifiability condition of testing and scientific status Severe test which, if the theory passes, the T may be said to be corroborated Rejection of the Duhem-Quine Thesis • Popper was anti-justificationist: o Justificationism - If there is a reason for believing in something, then we have reasons to believe that the theory is true o Anti-justificationism • Necessary conditions for falsification: o Popper held that a single falsification will not work by itself but the falsification must be repeatable o "We shall take a theory to be falsified only oif we discover a reproducible effect which refutes the theory" o Only accept refutations of theories from lower level hypotheses or generalisations describing reproducible effects • Nola holds a Theory is testable if it has a testable consequence

• (1) Fresnel Poisson Spot prediction • Newton - particle theory. • Fresnel provides a new kind of wave theory and presents result to French Academy (1818) • Poisson works out maths for a circular opaque disc and says that there would be a bright spot in the middle - which he thought impossible. • Arago does the experiment - and there is a bright spot! Calls it "Poisson spot"! • Franceso Grimaldi noted this more than 100 years before - but his paper was forgotten • (2) Einstein bending of light prediction from GTR • Huygens - first wave theory of light

Popper's Aim for Science: • FRRMV 1. Falsifiability 2. Critical revisability a. Distinguishing characteristic of empirical or scientific claims are their 'susceptibility to revision' b. Contra: religious beliefs etc. c. Two ways of revising beliefs i. Thinking ii. Observation iii. (Nothing like intuition, revelation or faith) 3. Critical rationalism a. For all rational discussion, not just philosophy or science, is 'stating one's problems clearly and of examining its various proposed solutions critically. b. "whenever we propose a solution to a problem, we ought to try as hard as we can to overthrow our solution to a problem rather than defend it. Few of us, unfortunately, practice this precept.' P, LSD, p 18 4. Maximise falsifiability a. Explanatory depth follows from theories which are maximally falsifiable - The more it explains, the more falsifiable it is 5. Aim to get highest verisimilitude or truth likeness a. Minimise false consequences and maximise true ones • Contra other aims o To get Certainty HD theories do not get at certainty Almost no 20th Century philosopher of science proposes certainty as an aim of science o To verify It cannot, according to P, be realised o To get truth Popper thinks it is highly unlikely that you can get at the truth of anything Replaces it with the idea of verisimilitude - the truths they capture grow as time goes on. Their truth likeness increases gradually and don't automatically become true. We will aim for the truth but won't ever get there Aim to show that one theory is more like the truth than the other

What is Popper's demarcation and model of science? Plus other respectable things. • Demarcation o Admit into science only those statements that a testable (i.e. only falsifiable) o and reject all else o Add rules below • Popper's model of science o Definition of testability or scientific theory o His aims o Demarcation • Other respectable things o Mathematics o Logic

What are Popper's Rules, the logical form of some Ad Hoc stratagems and the way the rules can be tested? • Described as conventions 1. Anti-dogmatist a. "The game of science is, in principle, without end." b. Maybe it means, don't give up testing, especially when new test situations arise 2. Rejection and acceptance rules a. Reject theories if they are falsified or less probable than a superior theory 3. Meta-rule: a. Rules must be designed in such a way that they do not protect any statement in science against falsification • Popper's Anti-Ad Hoc Rule 1. Latin: Ad hoc - 'for this' 2. For this purpose, or this case. only. 3. Ad Hoc committee - to do just this 4. Ad hoc solution to a problem; a supposed solution which is not generalisable; tailor-made for the one problem case. • 'As regards auxiliary hypotheses we decide to lay down the rule that only those are acceptable whose introduction does not diminish the degree of falsifiability or testability of the system in question, but on the contrary, increase it'. (LSD pp. 82-3). • Logical structure of some ad hoc stratagems: • T →P; • But P is false; So T is false (By Modus Tollens) • Introduce saving hypothesis or ad hoc hypothesis A such that (T&A) do not entail P, but avoids P: • T(modified) * P • Attack on Popper's Rule: Paul Feyerabend, Against Method • Some non-scientific examples • Rugby • H = No NZer hates Rugby. • I do! • H(modified) = No decedent Kiwi joker hates Rugby • Contraction of antecedent to make claim unfalsifiable in face of counter-instance? • Swans • H = All swans are white • Observe Australian Black Swan • H(mod) = All European swans are white. • Or; H (mod*) = All real swans are white. • Scottish people • H = All Scottish people love to eat haggis; • Hamish and Fiona hate haggis; • H* = All true Scottish people love to eat haggis. • Pacifist mediums • H = No democracy has every declared a war on another country. • NZ declared war in Korea, Vietnam, etc. • H* = No true, or mature, democracy has every declared war on another country. • Mediums • H = All mediums are able to contact the dead on the "other side"; • Medium X was not able to contact my dead relatives; • H* = All mediums not in the company of hostile disbelievers are able to contact the dead. • The problem of Job (Bible) • God is all-powerful, all-knowing and all-benevolent • Job: I have suffered all my life. • God is all-powerful, so could have prevent it. • God is all-knowing, so does not fail to know about my suffering; • God is all-good, so how can he let suffering go on? • So, such a God does not exist. [Argument from Evil] • Saver Modification: God does exist. & Your suffering is visited upon you as a test of your faith in his existence. • Q: is the modification a saver of God Hypothesis? • Q: is the saver modification testable? • Two kinds of ad hoc • 1: Revise theory and assumptions to avoid false prediction with no new testable hypotheses • 2: Revise theory and assumptions and add testable predictions which are false • Ad-hoc rules • Revision against false predictions must have extra testability or more predictions • Revision concerning false evidence to the predictions • Only can discredit the data when there are counter-observations or counter-experiments • Theories should not be definitions (which are unfalsifiable) but treated as empirical claims • If we change meanings in theories, then such changes should add to the overall falsifiability of the hypothesis • Testing conventionally adopted rules • Intuition test - Popper initially proposes that the rules be tested against the intuitions of scientists as to what science is • Science test - Popper changed this strategy and later drew up a list of the good and bad science - Einstein newton Darwin on one side and freud and marx on the other - and saw whether the rule accounted for the good and bad science - requires some intuition though

Ad hoc case studies • Case Study • Leon Festinger: Cognitive Dissonance • Leon Festinger, Henry Riecken, and Stanley Schachter, When Prophecy Fails (1956) • Marian Keech (real name Dorothy Martin), a follower of Hubbard Dianetics/Scientology, forms new small group. • Received automatic writing (from inhabitants of planet Clarion) saying that world would be destroyed at 7 am, 21 December 1954; but members of her group would be rescued in a flying saucer at midnight just before destruction. • Prior to 20 Dec group shuns publicity and remains secretive (but Festinger infiltrates it). • On 20 Dec groups assembles, removes all metallic objects from themselves (as instructed) and wait for UFO. • Midnight - nothing happens! • Clock in another room says 11.55 pm - so adjust time. • Festinger Cog. Dis. continued • 12.10 am. Supposed midnight and no UFO with less than 7 hours to cataclysm! Prediction refuted?? • 4 am. No satisfactory explanation - Keech cries! • 4.45 am. More automatic writing arrives. God decides to spare Earth from destruction: "The little group, sitting all night long, had spread so much light that God had saves the world from destruction." • Afternoon of 21 Dec. Newspapers are called; interviews are sought. Reversal of its previous dislike of publicity. Group begins campaign to spread its message to as broad an audience as possible. No longer secretive. • Later witch-hunt to seek out those who are not true believers so group becomes more "pure". • Falsified prediction leads to cognitive dissonance and various attempts to relieve it. • Cognitive dissonance between what one believes and what actually happens • Case study 2: Phlogiston Theory • Theory of combustion before oxygen theory: • H = All substances (metals) when they burn give off phlogiston • Conflicting discovery that burnt residue is heaver after combustion than before. • So, (H&A) = phlogiston has negative weight • Or (H&A*) = an unknown substance joins the burning material, which has more weight than the phlogiston that leaves it. • Case study 3: Perfect heavenly bodies • Aristotelian H = all heavenly bodies are perfect spheres. • Galileo: Observed with a telescope that the Moon's surface was pitted and cratered. • Aristotelian modification (H&A*) = Craters and pits on Moon are filled with an invisible-to-us substance. • (Q: How high is it piled up?) • A. Chalmers WITTCS p 51. • G. Hempel, Philosophy of Natural Science chapter 3 • Stellar parallax: Earth moving 1→2 makes it appear as if star moves wrt background • Stellar parallax • Copernicus/Galileo H = Earth revolves around the Sun. • ARISTOTLE: If Earth rotates around the Sun then we should observe stellar parallax. • We do not observe it, • So the Earth does not rotate about the Sun. • Copernicus/Galileo modification (H&A) = Earth revolves about Sun; but are so far away from the stars that we do not observe it • First observed 1838: Friedrich Bessel's parallax of 61 Cygni • Horror Vacuui • Nature abhors a vacuum. • H = The stuff of nature will always rush in to fill any gap of nothingness in nature. • Torricelli apparatus taken up a mountain; observed that the gap at the top of the tube increased; i.e., the mercury column fell • Saving hypothesis (H&A) = the strength of the abhorrence of a vacuum varies with height. • G. Hempel, Philosophy of Natural Science p 28-9. • Optimality Hypothesis • H = All creatures evolve to have a maximum number of offspring. • Optimal behaviour of gannet: all gannets normally lay one egg. • Counter-observation: when an extras egg is put in a gannet nest normally the adults can rear both chicks. In other words, they can have two eggs but they just lay one. • Saving "Time Lag" hypothesis (H&A) = There is a time lag during which gannets adjust their optimal behaviour wrt clutch size. • Gannets are "out of date" - genes selected to build them to function at 'some earlier era when conditions were different' • R Dawkins, The Extended Phenotype, p. 35. • Motion of the Earth. • Aristotelians: Earth does not move (rotate on axis) • Galilean H = Earth does move (rotate on axis) • Aristotelian objection; Why are birds still with us? • Tower Experiment: Stone dropped from top of tower falls at bottom of tower, not away from bottom • Galileo reply (H&A) = All bodies (birds before flight, stones at top of tower) share in the Earth's rotation. • How would you test Galileo's reply? Is the sharing detectable by us on Earth? (Foucault's pendulum?) • Galileo's Tower Experiment • 1: Aristotelians • Stone falls to base • ↓ (proves) • Earth at rest • Motion of Earth • ↓ (Predict) • Stone falls away from base (false prediction) • 2: Galileo and Relative Motion • Stone falls to base • ↓ (proves) • No relative motion between stone and Earth • Motion of Earth • ↓ (proves) • No relative motion between system of stone, tower and Earth • First piece of evidence to distinguish • Tilting an umbrella in rain • Tilting the mircroscope • Foucault's pendulum where the pendulum swung in a different direction from a dome in which it started. • Anomalous behaviour of orbit of Uranus. • Uranus first planet to be discovered in modern times - John Herschel in 1781. (observed before but mistaken for a star) • Using Newtonian mechanics Laplace calculated its orbit in 1783; but soon after discrepancies between calculated and observed orbit were noted. • Was Newtonian Mechanics wrong? • Was there a perturbing planet outside Uranus? • In 1840s John Adams and Urbain Joseph Leverrier investigated the discrepancies and suggested where one might find a perturbing outer planet. • In 1846 Johann Galle reported a planet close to where Leverrier predicted it was. Named 'Neptune'. • Saving hypothesis of extra planet that vindicated Newton's theory. • See paper by N. R. Hanson on this "Zenith & Nadir". • Precession of Perihelion of Mercury. • Perihelion - peri, closest point on orbit to, Helios Sun • The perihelion of Mercury precesses -point slowly shifts with each orbit of the Sun. Elliptic orbit of Mercury slowly rotates. • Total amount (Urbain Joseph Leverrier in 1840s): 574 arc-seconds per century. (in 2,250 centuries complete rotation) • Perturbation due to Venus: 277 arc-sec • Perturbation due to Jupiter: 153 arc-sec • Perturbation due to Earth: 90 arc-sec • Perturbation due to other planets: ≈10 arc-sec • Total perturbatory effects about 531 arc-sec/century • Missing 43 arc-sec/century! A crisis-causing anomaly for Newtonian mechanics! • Saving Hypotheses • There is a planet inside Mercury - 'Vulcan'. Observed? • There is a Vulcan planet inside Mercury which accounts for the anomaly but cannot be seen from Earth since it has same period as Earth (not physically possible system). • There is matter, ring of particles or "asteroids", inside Mercury that causes anomaly • Simon Newcomb (1890s): Modify Newton's laws. Inverse square law becomes inverse 2.000,000,1574. Accounts for anomaly! • Sun is not a perfect sphere and is oblate; so irregular gravitational field for Mercury. • Einstein's General Theory of Relativity (1915) explains exact missing 43 arc-sec/century. [Known non-designer evidence] • In 1960s Dicke, Goldenberg and Brans revive solar oblateness theory to account for about 3-4 arc-secs • See George Will, Was Einstein Right? Chapter 5.

What are Conventionalists? - Constructivism comes with conventionalism - According to Popper, conventionalism includes the following: - Theories as definitional conventions - o This rejects the demarcation and ad hoc rule 3 o Properties of the world do not determine the theory but rather the theory determines the properties - Ad hoc hypotheses admissible o Any hypothesis can be introduced to save a theory - Conventionalist rules for theory acceptance/rejection where things are equal o Whether it uses less or more ad hoc hypotheses o Complexity in the structure of its implicit definitions o Simplicity (unspecified though) o Greater explanatory power o Number of basic principles is simpler in respect of number of independent assumptions o Elegance and aesthetically pleasing

What is Popper's Response to Conventionalism? • It is consistent and defensible • The only way to avoid conventionalism is to decide not to accept it • Popper could have o Said conventionalism doesn't provide any knowledge or insight into nature o Doesn't mould with the intuitions of scientists

• Survival of the fittest was used by Darwin in later versions of the Origin of species but is by Huxley - Nola thinks it reduces his views a bit. • Facts: • Fact: Huge variety of species. Fact noted by biologists before Darwin. Why such diversity? • Fact: Species marvellously adapted to their Environment on land, sea and/or air. • Fact: phenotypic similarities within species (humans) and between species, i.e. homologies. Why such similarity? • Explanations in terms of essential nature?? • Problem of intermediate cases: marsupials; platypus, egg-laying, duck-billed mammal. • Problem of what fossils are and how they came about. • Darwin's OS: "one long argument for evolution".

• Outline: • Step 1 - Fact of Variation/Diversity • Step 2 - The "crunch point" of existence • Step 3 - Utility of the Variations • Step 4 - Strong Inheritance Principle (SIP) • Argument • Step 1- Fact of Variation/Diversity and advantages • Variation • Great variation in the characteristics of organisms within a species, i.e., their phenotype. • Variation between individuals in a species. • Advantages: • All of above known by breeders through Domestic selection. Select race horses for speed; select cows for milk production, select chickens for # eggs laid, • Step 2 - The "crunch point" of existence • Different species have different niches in same environment. Finches eat seed, thrushes worms, etc • For a species S in its niche, the Environment can sustain only so many members of S. • If S not at "crunch point", then no extreme competition within species S. • If S at or near "crunch point" then competition between members of S is extreme. • Doctrine from Thomas Malthus on population • Thomas Malthus 1766-1834 • Essay on Population 1798 • Prediction that population size would outrun food supply - massive die-off of humans. (not correct, yet) • Population increases geometrically 2, 4, 8, 16, 32, 64, • Food production increases arithmetically, 2, 3, 4, 5, 6, • Natural causes will restrict population (war, disease, famine, climate). But not by restraint, moral or other. • In nature animals & plants have more off-spring than actually survive. Same for humans at time of Malthus • Malthus' thesis: there is a maximum number of members of a species that any given environment can sustain. • There is a 'crunch point' of competition for survival. • Darwin on Influence of Malthus • In October 1838, that is, fifteen months after I had begun my systematic inquiry, I happened to read for amusement Malthus on Population, and being well prepared to appreciate the struggle for existence which everywhere goes on from long-continued observation of the habits of animals and plants, it at once struck me that under these circumstances favourable variations would tend to be preserved, and unfavourable ones to be destroyed. The results of this would be the formation of a new species. Here, then I had at last got a theory by which to work". (1876 Autobiography) • The observations didn't give him his theory, but Malthus did • He was a hypothetico-deductivist, not a inductivist • Step 3 - Utility of the Variations • Given observed fact of variation of phenotypic characteristics of members within a species 1. Some variations confer advantage 2. Some variations confer disadvantage 3. Some variations are neutral. • 1, 2, & 3 have been observed by naturalists. • Some characteristics make some members of a species better, or less, able to survive in their environment. Environment may change making some characteristics advantageous or disadvantageous. • Step 4 - Strong Inheritance Principle (SIP) • Off-spring are more like their parents than other members of the species. • Darwin knew nothing of genetics. • But all people know that "something" is passed from parents to off-spring, and they have their parents phenotypic characteristics. • SIP is an obvious claim that all assent to. • Useful to add what we now know about genetics to what Darwin says of NS. Get Modern "synthesis" of Gen+Ev. • Argument for Evolution by NS: Step 5 • Consider above 4 claims: 1. Great phenotypic diversity, variation, within species. 2. Often species in their Environment are at, or close to, Malthusian "crunch point". 3. Variations are advantageous or disadvantageous. 4. Strong Inheritance Principle 5. Advantageous variations survive down generations 5. Disadvantageous variations do not survive down the generations (Members with disadvantageous variations either die or have no surviving off-spring) • Frequency of many phenotypes alters down the generations. That is, species change, evolve in their phenotypic characteristics. NO essence here! • Bacteria have evolved and survived anti-bodies and anti-biotics. • Bacteria have evolved and survived anti-bodies and anti-biotics.

• Outline • LOL • Tree of life • Four of many cases: • Homology - morphological and genetic • Darwin non-aquatic or non-web-footed birds • Palaeontology - Transitional forms without leaps • Blind cave insects • Elliott Sober: Law of Likelihood • Sober, Evidence & Evolution, p 32, section 1.3 • LoL: Observations O [Facts O] favour hypothesis H1 over hypothesis H2 = [if and only if] prob(O, H1) > Prob(O, H2). And the degree to which O is favoured is given by the likelihood ratio prob(O, H1)/Prob(O, H2). • Different from IBE in which one argues from better explanation of O by H1 than by H2 to truth, or acceptance, or H1 over H2 • Favouring is a different, weaker notion. Sober thinks of it as differential support. • Darwinian Tree Structure of Evolution • For every species S (except the first) there is an immediate ancestor species A such that S evolved from A. • This forms a tree structure of relations between species, from distant ancestor to many descendants. • Case 1: Homology • Humerous, ulna, radius etc. are common in frog, lizard, bird, human, cat, whale and bat. • Different Function - yet Homologous? • What can be more curious than that the hand of man, formed for grasping, the mole for digging, the leg of the horse, paddle of the porpoise, and the wing of a bat, should all be constructed on the same pattern, and should include the same bones, in the same relative positions? .... the parts may change to almost any extent in form and size yet they always remain connected together in the same order. We never find, for instance, the bones of the arm and forearm, or of the thigh and leg, transposed. .... • Nothing can be more hopeless than to attempt to explain this similarity of pattern in members of the same class by utility or the doctrine of final causes [teleology, purpose]. ... On the ordinary view of the independent creation of each being, we can only say that so it is: - that it so pleased the Creator to construct each animal and plant. (Darwin OS 415-6) • Homology: Surprise or no Surprise? • Case of four limbs • Whales with useless hind limbs 1) In NS, the many facts of homology is no surprise. • Evolution by NS from a common ancestor - homology is to be expected. • prob(the various classes of fact about homology between pairs of species, given that NS) = high 2) For Creationism (CR) many facts of homology are a surprise. On CR, homology is unexpected. • Because there are many other possibilities for Creator other than the few of homology, or structure. • prob(the various classes of fact about homology between pairs of species, CR) = low, or zero. 3) Consilience of facts of homology by NS but no consilience by CR • Creationism leaves the separate facts of homologies unaccounted for but NS pulls them together 4) NS makes facts of homology more probable than CR. 5) Likelihood of NS greater than CR. • Homology: The Grand Genetic Homology • Millions of alternative, possible genetic codes exist (1.5x10 power 84!); yet nearly all organisms have the same genetic code (small differences - mitochondria). • Genetic code based on chemistry of DNA molecule • The genetic code is homologous for (most) organisms. • Why? What is the best explanation of this? • The hypothesis of evolution is a good explanation. • The common ancestor to all known organisms had a genetic code similar to what we find today. Over the ages, the genetic code has passed unchanged (or nearly so) from parents to offspring. Given NS, homology expected. (Darwin on homology in embryonic forms - Origin of Species chapter 13) • Creationist explanation: if all creatures were created separately, why should we expect that they have homologous genetic codes? Not expected at all. • prob(Homology, NS) > prob(Homology, CR) • Likelihood of NS greater than CR • Case 2: Darwin on birds • He who believes that each being has been created as we now see it, must occasionally have felt surprise when he has met with an animal having habits and structure not at all in agreement. What can be plainer than that the webbed feet of ducks and geese are formed for swimming? yet there are upland geese with webbed feet which rarely or never go near the water; and no one except Audubon has seen the frigate-bird, which has all its four toes webbed, alight on the surface of the sea. On the other hand, grebes and coots are eminently aquatic, although their toes are only bordered by membrane. ... In such cases, and many others could be given, habits have changed without a corresponding change of structure. The webbed feet of the upland goose may be said to have become rudimentary in function, though not in structure. In the frigate-bird, the deeply-scooped membrane between the toes shows that structure has begun to change. (OS 185-6: Chapter VI, Difficulties of the Theory) • Darwin on web-footed birds • Habits of birds change but not structure. Have webbed feet but do not go near water; have no (hardly any) webbed feet but live off water. 1) prob(web-footed birds which are not aquatic, CR) = low 2) prob(aquatic birds that are hardly or not web-footed, CR) = low 3) prob(web-footed birds which are not aquatic, NS) = high 4) prob(aquatic birds that are hardly web-footed, NS) = high • Explain facts: (1) webbed feet but do not live off water; (2) live off water but no webbed feet; • Both explained by NS but not well explained by CR • Additional Comments by Darwin • He who believes in separate and innumerable acts of creation will say, that in these cases it has pleased the Creator to cause a being of one type to take the place of one of another type; but this seems to me only restating the fact in dignified language. He who believes in the struggle for existence and in the principle of natural selection, will acknowledge that every organic being is constantly endeavouring to increase in numbers; and that if any one being vary ever so little, either in habits or structure, and thus gain an advantage over some other inhabitant of the country, it will seize on the place of that inhabitant, however different it may be from its own place. Hence it will cause him no surprise that there should be geese and frigate-birds with webbed feet, either living on the dry land or most rarely alighting on the water; that there should be long-toed corncrakes living in meadows instead of in swamps; that there should be woodpeckers where not a tree grows; that there should be diving thrushes, and petrels with the habits of auks. (OS 185-6: Chapter VI, Difficulties of the Theory) • Case 3: Transitional forms without Leaps • Although in many cases it is most difficult to conjecture by what transitions an organ could have arrived at its present state; yet, considering that the proportion of living and known forms to the extinct and unknown is very small, I have been astonished how rarely an organ can be named, towards which no transitional grade is known to lead. The truth of this remark is indeed shown by that old canon in natural history of "Natura non facit saltum." We meet with this admission in the writings of almost every experienced naturalist; or, as Milne Edwards has well expressed it, nature is prodigal in variety, but niggard in innovation. Why, on the theory of Creation, should this be so? Why should all the parts and organs of many independent beings, each supposed to have been separately created for its proper place in nature, be so invariably linked together by graduated steps? Why should not Nature have taken a leap from structure to structure? On the theory of natural selection, we can clearly understand why she should not; for natural selection can act only by taking advantage of slight successive variations; she can never take a leap, but must advance by the shortest and slowest steps. (OS 194: Chapter VI, Difficulties of the Theory) • Probability of transitional forms without Leaps 1) prob(there are transitional forms for each organ and no leaps of structure, NS) = high 3) prob(there are transitional forms for each organ and no leaps of structure, CR) = low 4) for the FACTS of transitional forms with no leaps of structure, (1) & (3) better explained by (not surprising) NS; not explained, and are surprising, by CR. • We have in this chapter discussed some of the difficulties and objections which may be urged against my theory. Many of them are grave; but I think that in the discussion light has been thrown on several facts, which on the theory of independent acts of creation are utterly obscure. (OS 203) • Says that, although he has some problems for his theory, creationism is worst of • Case 4: Darwin on blind insects in caves • It is difficult to imagine conditions of life more similar than deep limestone caverns under a nearly similar climate; so that on the common view of the blind animals having been separately created for the American and European caverns, close similarity in their organisation and affinities might have been expected; but, as Schiödte and others have remarked, this is not the case, and the cave-insects of the two continents are not more closely allied than might have been anticipated from the general resemblance of the other inhabitants of North America and Europe. On my view we must suppose that American animals, having ordinary powers of vision, slowly migrated by successive generations from the outer world into the deeper and deeper recesses of the Kentucky caves, as did European animals into the caves of Europe. We have some evidence of this gradation of habit; for, as Schiödte remarks, "animals not far remote from ordinary forms, prepare the transition from light to darkness. Next follow those that are constructed for twilight; and, last of all, those destined for total darkness." By the time that an animal had reached, after numberless generations, the deepest recesses, disuse will on this view have more or less perfectly obliterated its eyes, and natural selection will often have effected other changes, such as an increase in the length of the antennæ or palpi, as a compensation for blindness. OS 138: Chapter V Laws of Variation) • Darwin on blind insects in caves 1) prob(same design for blind insects in similar caves in America and Europe, CR) is high; 2) prob(different design for blind insects in similar caves in America and Europe, CR) is low. 3) prob(internal and external insects are similar, CR) = prob(internal and external insects are dissimilar, CR) = ½. 4) prob(blind insects in caves are similar to (some) insects external to caves, CR) ≤ ½. 5) prob(FACT1: different design for blind insects in similar caves in America and Europe, NS) is high. 6) prob(FACT2: blind insects in caves are similar to (some) insects external to caves, NS) is high 7) So, NS is better explainer than CR

What does Darwin say about creationism against the evolution facts and what is the problem of Gods wants? • Darwin: CR offers no explanation • "He who believes in separate and innumerable acts of creation will say, that in these cases it has pleased the Creator to cause a being of one type to take the place of another type; but this seems to me only restating the fact in dignified language." (Darwin OS 217) • "On the ordinary view of the independent creation of each being, we can only say that so it is: - that it so pleased the Creator to construct each animal and plant." (Darwin OS 416) • Problem of getting evidence for what Creator desires or wants. Evidence impossible to get? • God Hypothesis is Useless Explanation • Human model of explanation appeals to our beliefs and desires (wants, goal, ends) • Extended to God - same belief-desire explanation. A personal God? - God has aspects of a person • Supposition: God is omniscient - has all true beliefs and no false beliefs • God's wants, desires - can we know what they are? • Is there Impiety in supposing we can read God's mind! • Two problems with God's wants: 1. Means whereby God's wants are realised? Either God is part of the naturalistic world and causally interacts with it in the familiar way; or God is not part of the natural world and we need to invent a special notion of Divine-Causality to explain his interaction with it. 2. Which wants does God have? Unknowable. Impious? • There are upland geese with webbed feet that do not go near the water. Did God want this? Evidence? Or just "God wanted it and that is that!" • Is there any evidence that God wanted non-aquatic web-footed birds, N-AWFB, independent of the fact of N-AWFB? • Argument: 1. There are N-AWFB 2. God has the power to create N-AWFBs [grant omnipotence] 3. If God has power to create N-AWFBs (without bungling!), and there are N-AWFB, then God must have wanted N-AWFB. 4. Therefore: God wanted that there be N-AWFBs • 3 begs issue of Creationism! Is 3 true? • No independent evidence for what Gods wants. If there are N-AWFBs then God must have wanted them! • God makes Each Separate • Fact of Homology • God's Decision #1 Human Arm Structure • God's Decision #2 Lizard Limb Structure • God's Decision #3 Whale Flipper Structure • God's Decision #4 Cow Foreleg Structure • God's Decision #n Bat Wing Structure • William Whewell again • Whewell epigraph at beginning The Origin of Species • "But with regard to the material world we can at least go far as this - we can perceive that events are brought about not by the insulated interpositions of Divine power, exerted in each particular case, but by the establishment of general laws." • SC: World is an ocean of separate facts made by a Creator. • Lacks any underlying natural mechanism, cause or laws. Creator just so wills. Is there then no science apart from noting any observable regularities? • Contrast with NS: there are natural mechanisms and laws at work to be discovered; so science is possible. • God makes only World + Laws of Nature - Leaves World to run by itself [adapted from Elliott Sober E&E chapter 2] • Perhaps: • 1. God's one decision of the world • 2. World laws of nature and then this leads to • 3. Below facts • Human Arm • Lizard Limb • Whale Flipper • Cow Foreleg • Bat Wing • Darwin on God versus Science • "Astronomers might formerly have said that God ordered each planet to move in its particular destiny. In same manner God orders each animal created with certain form in certain country, but how much more simple and sublime power let attraction act according to certain law, such are inevitable consequences - let animal be created, then by the fixed laws of generation, such will be their successors. [let the theory of attraction explain how the planets move] Let the powers of transportal be such, and so will be the forms of one country to another.- Let geological changes go at such a rate, so will be the number and distribution of the species!!" [Darwin, 1837-39: Notebook B] • '. . . astronomers do not state that God directs the course of each comet and planet. The view that each variation has been providentially arranged seems to me to make Natural Selection entirely superfluous, and indeed takes the whole case of the appearance of new species out of the range of science' [Letter] [Not postulating the mechanisms - why can't you be a creationist about everything]

What is Abduction, some examples and forms, a fallacy, an account of the BE, final form of abduction, comments on conclusions and consilience? • Example of IBE 1. Fact F = all student essays the same • Hypothesis H1 = students copied one essay • Hypothesis H2 = students worked independently 2. H1 explains F very well. Alternative accounts: F is no surprise given H1. Or, if H1 were true, expect F to be true. Or probability of F given H1 is high. Prob(F, H1) is high. 3. H2 explains F badly. Alternative accounts: • F would be a surprise given H2. Or if H2 were true expect F to be false. Or probability of F given H2 @ 0. Prob(F, H2) is low. • Hence name Surprise/No-Surprise principle. • [probability: likelihood ratio] • Inductive Generalisation as IBE o F1: Sun rose yesterday; o F2: Sun rose 2 days ago; o F3: Sun rose 3 days ago; o ..................................... o Fn: Sun rose n days ago o F = F1 & F2 & F3 & .... Fn. o Explanatory Hyp. H1: Sun rises every day o Explanatory Hyp. H2: Sun will not rise tomorrow o Explanatory Hyp. H3: ???? o H1 explains F better than H2, or any H3? o Is there a better explanation than H1 that you can envisage? • "only game in town fallacy" 1. Fact: odd noises in attic during night 2. H1 = Martians have landed and are playing bowls in attic 3. Have no other hypothesis to suggest. 4. If H1 true then F is no surprise. 5. Accept H1 as true. • But H1 is the "only game in town". American colloquialism for it's the only hypothesis • H1 is implausible given what else we know. • An Account of Best Explanation • Best Explanation = Maximum Likelihood Plus High Prior Probability • For two potential explanations H1 and H2, and some fact F and K is background beliefs other than F: • H1 is a better explanation of F than H2 =Defn • Likelihood (i) prob(F, H1) > prob(F, H2), • Prior Probability (ii) prob(H1, K) > prob(H2, K). Or at least probK(H1) = probK(H2) • On the absence of prior probability information, make the likelihoods equal • Proposal: We are to understand Darwin's claim that NS is a better explanation than CR = • (i) given fact F, NS has higher likelihood than CR; • (ii) prior probability of H1 either equal to, or greater than, H2. • Final Form of IBE/Surprise Principle 1. F fact(s); 2. H1 explains F to some degree d; 3. Rival H2 explains F to some degree d*; 4. Both H1 and H2 pass threshold for satisfactory explanations; (d, d* > minimum) 5. d is higher than d*; 6. So, H1 explains better than H2; 7. Accept H1 as true, rather than H2 as true • Conclusion: accept best explainer. Hence IBE • Note: If no rival H2 available, not-H1 is available. • Conditions of IBE: o Need two or more hypotheses o Need both to be minimally satisfactory explanations of F o Each must be plausible given what else we know. If K is what else we know other than F, then prob(H1,K) > prob(H2,K). K is background or "prior" information other than F. • Conclusions of Abduction • T is true - held by few • It is reasonable to accept T as true - • T is to be accepted over T* • T is to be favoured over T* - Some suggest it is Inference to the Most Favoured Hypothesis rather than IBE • Consilience - William Whewell (1794-1866) • Consilience - Theory jumps together distinct classes of fact, or of established laws and receives confirmation 1. Example: 2. Kepler's 3 planetary laws 3. K3 → Newton's Law Universal G • Consilience/Unity achieved for disparate K1, K2, K3 in light of Newton's Theory • Theory gets extra support by unifying/consiliating disparate facts

What is Paley's watch argument and Hume's critique? • Structure: 1. Paley's watch argrument 2. Hume's argument 3. Hume's counter-analogies 4. Form of inductive analogy 5. Good and bad inductive analogies 6. Objection from poor design and natural selection • NS IBE for BD • §2: Paley's Watch Argument 1) Watches are complex, intricate, law-governed, and were made by intelligent designers (watchmakers) C, I, L 2) The world is complex, intricate at all levels and is law governed; 3) So, by analogy, World also has its super-intelligent designers - supreme watchmakers Objections: M - One designer or a committee?? U - Argument to a unique God? I - Good analogy? Are watches and World sufficiently similar and dissimilar in right ways? • §3: Hume's Design Argument • "Look around the world: Contemplate the whole and every part of it: You will find it to be nothing but one great machine, subdivided into an infinite number of lesser machines... All these various machines, and even their most minute parts, are adjusted to each other with an accuracy, which ravishes into admiration all men, who have ever contemplated them. The curious adapting of means to ends, exceeds the productions of human contrivance; of human design, thought, wisdom, and intelligence. Since, therefore the effects resemble each other, we are led to infer, by all the rules of analogy, that the causes also resemble; and that the Author of nature is somewhat similar to the mind of man; though possessed of much larger faculties, proportioned to the grandeur of the work, which he has executed. By this argument a posteriori, and by this argument alone, we do prove at once the existence of a Deity, and his similarity to human mind and intelligence." (Dialogues Concerning Natural Religion Part 2, 'Cleanthes') • Structure of Hume's Argument 1. The world contains many complex artefacts, A, such as sundials, computers, cars, aeroplanes, etc; and these are made by intelligent designers (us!). 2. The natural world is full of complex "machines" M such as DNA, flowers, crystals, viruses, human bodies (hearts, brains) etc; 3. Ms resemble As; but Ms exceed As in their functions, complexity, interconnection, etc. 4. A's are caused to exist by humans with minds 5. So, by analogy, M's are also caused to exist by an entity with a mind like ours but much grander. 6. So, a Grand Intelligent Designer exists! • Spiders - Hume's Counter-Analogies: 1 1) Webs are complex, intricate, law-governed, and were made by intelligent designers - spiders 2) The world is complex, intricate at all levels and is law governed; 3) So, by analogy, World also has its super-intelligent designers - super-spiders Good analogy? Are Webs and World sufficiently similar and dissimilar in right ways? • Vegetables - Hume's Counter-Analogies: 2 1) Plants are parts of the world which are complex and intricate, law-governed; and they propagate by seeds. 2) The world is complex, intricate at all levels and is law governed; 3) So, by analogy, World also has its Great Vegetable in the Sky, which propagates the World by seed Good analogy? How different from the rest? • Form of Inductive Analogy 1) X has A, B, C, .., T; X is also Z 2) Y has A, B, C, .., T; 3) So, Y has Z Inductive Argument by Analogy How good are such arguments? They are as strong or as weak as the relevant similarities and dissimilarities are weighted against one another • Good and Bad inductive analogy • Flu • Person A shivers, sneezes, has aching joints, high temperature; and A has the 'flu. • Person B shivers, sneezes, has aching joints, high temperature • B has the 'flu. • Petrol • Aeroplanes are noisy, coloured and fly; and they run on petrol • Birds are noisy, coloured and fly • Birds run on petrol. • Degree of support = degree of overall similarity. Aeroplanes & birds share few features. • Person A and B have similarities and dissimilarities • Birds and planes have a much lower degree of similarities • Paley's Argument as Inductive Analogy 1) Watches are A & B & C &.. &Q, and were made by an ID-er (Intelligent Designer - watchmaker) 2) The world is A & B & C &.. &Q; 3) So, by analogy, World also has its ID-er (super-intelligent designer - supreme "watchmaker") • How strong is the analogy? What are the similarities and dissimilarities between watch and world? Do dissimilarities outweigh similarities? So analogy is weak? World has features F1 to F(million) that watches do not have? • Paley envisages his watches giving birth to one another so as to make them more like species. Does this help and argument to an ID-er of species if not the world? • Argument from Poor Design 1) If Creator, or Intelligent Designer, [God], is omnipotent, omniscient and omni-benevolent (OOO-God) and creates organisms, then it would create organisms that have features that exhibit optimal design; 2) Organisms have features that are not optimal. 3) Either OOO-ID-er [God] did not create these suboptimal organisms; or ID-er is not omnipotent, omniscient and omnibenevolent. • Valid? [Modus Tollens & De Morgan law] • Premises true? Auxiliary to premise 1: could OOO-Creator have an intention to make suboptimal creatures? Creator as malevolent or as a bungler? • NS or SC better explainer of Poor Design? • Poor design can be expected by theory of NS. • Features that were evolved for certain uses, are then reused or co-opted for different uses, or abandoned altogether (appendix, female birth canal, Huntington's Disease (CAG repeats)) • Sub-optimality is due to the inability of the evolutionary process to eliminate the particular vestiges of earlier aspects of the evolutionary process. • Can poor design be expected on basis of Special Creationism? • Using IBE, fact of poor design favours NS over SC

• Elliott Sober: Alternative Form of Paley's Argument as IBE • Looked at Paley's Argument as an Inductive Analogy. • Somewhat inconclusive. • Alternative interpretation as Inference to the Best Explanation (IBE) • Or the Surprise Argument • §4.1: Abduction: Inference to the Best Explanation (IBE): Surprise Principle 1. Suppose we are given some facts F. 2. Suppose also given rival theories or laws: A and B. 3. A explain F better than B. • Question: Is being a better explanation a guide to the truth (or to increased verisimilitude)?? • A explains better, so is "more true" than B? If so ... 4. A is to be accepted over B (?) • Here two values Explanation and Truth are allegedly linked. (This is a disputed matter!) • Inference from premises1 & 2 & 3 to conclusion (4) is not deductive; special case of induction. • Example of IBE 1. Fact F = all student essays the same • Hypothesis A = students copied one essay • Hypothesis B = students worked independently 2. A explains F [How well does it explain?] 3. B explains F [How well does it explain?] 4. A is a better explanation of F than B. 5. Accept A over B; or accept A as true. C. S. Peirce: Abduction - in contrast to both deduction & induction. Inference to the truth of the best explainer. • Further Analysis of IBE 1. Fact F = all student essays the same • Hypothesis A = students copied one essay • Hypothesis B = students worked independently 2. A explains F very well. Alternative accounts: • F is no surprise given A. Or, if A were true expect F to be true. Or probability of F given A is high, @ 1. 3. B explains F badly. Alternative accounts: • F would be a surprise given B. Or if B were true expect F to be false. Or probability of F given B @ 0. • Hence name Surprise/No-Surprise principle. • [probability: likelihood ratio] • Sober: "only game in town fallacy" 1. Fact: odd noises in attic during night 2. H1 = Martians have landed and are playing bowls in attic 3. Have no other hypothesis to suggest. 4. If H1 true then F is no surprise. 5. Accept H1 as true. • But H1 is the "only game in town". • H1 is implausible given what else we know. - Extra conditions on IBE argument: 1. Need two or more hypotheses 2. Need both to be minimally satisfactory explanations 3. Each must be plausible given what else we know. • Final Form of IBE/Surprise Principle 1. F is some (surprising ) fact(s) (of biology); 2. H1 explains F to some degree d; 3. Rival H2 explains F to some degree d*; 4. Both H1 and H2 pass threshold for satisfactory explanations; (d, d* > minimum) 5. d is higher than d*; 6. H1 has more initial plausibility than H2 (independently of F) 7. H1 explains better than H2; 8. Accept H1 as true, rather than H2 as true 9. Prefer or favour H1 over H2. • Form of argument is inductive - not deductive. • Conclusion: accept best explainer. Hence IBE • Note: If no rival H2 available, not-H1 is available. • §4.2: Application of IBE to Paley Argument 1. F = facts about order O, intricacy I, complexity C, law-governed character L, of all things = OICL 2. Hypothesis D = All OICL things have a designer 3. Hypothesis R = all things, including OICL things, come into existence Randomly 4. How probable (surprising) is F, given D? High? 5. How probable (surprising) is F, given R? Low? 6. Suppose 4 is High and 5 is Low. 7. So, D explains F better than R; 8. by IBE, Accept D as true, rather than R as true • Form of argument is inductive - not deductive. But not inductive Analogy - rather IBE

What is the definition of randomness and non-randomness and its application to evolution? • Random & Non-Random • Definition: A series of events is random = each event is as equally probable as any other in sequence of events. • Examples 1) Shuffle deck of 52 cards - pick one. Replace. Shuffle - pick again. Random? 2) Win a lottery of 100,000 tickets. Random choice of winner - how?. 3) Lotto - random? 4) Turn Roulette wheel? 5) Wind blowing leaves about a yard. 6) Tornado in junk yard assembles a Cadillac! • Non-Random Processes • A series of event is non-random = some events in the series are more likely than others. • Examples 1. Toss biased coin weighted 10 times to tails. 2. Length of life random/non-random? 3. Length of life given one eats chips at all meals • Evolutionary Natural Selection (NS) not Random; Sober Section 2.4 • Variation is random (although some physical factors may play a part), but natural selection is not random since not all the variations have an equal probability of being preserved • Need to examine: probability(S7, given S6), probability(S6, S5), etc • Evolutionary Hypothesis E: All species evolve from earlier species through a series of small steps by processes of natural and sexual selection • Random Hypothesis R: All species evolve as a whole from one of many immediately prior states which are equally probable. Examples: Cadillac from stuff in junk yard; molecular structure of body from heap of atoms. • Analogy as Example: • numbers 0 to 9 placed on a discs. Three discs next to one another. • Target number is: 267. Start Number is 265. • Rotate last disc: probability(267, given 265) = 1/10 • Target number if 267; Start number is 253. • Rotate 2 discs: probability(267, given 253) = 1/100. • Target number 267; start number is 593. Probability?

How does the evolution hypothesis relate to Paley's argument? • Outline: • Paley's argument • God hypothesis and tricky designer • Lamarck's false theory • Evolution as small steps or changes • Back to Paley's Argument 1) Facts: OICL facts possessed by most things 2) Hypothesis 1: D = Designer 3) Hypothesis 2: R = Random coming into existence 4) Hypothesis 3: E = NS by sequence of small steps. 5) Which of D, R, E explains F best? 6) Knock out R! 7) E explains F better, worse, than D? 8) Probability(F, D) and Probability(F, E) ??? 9) Paley's argument not conclusive against E. 10) Need further argument to show E is better explainer. Turn to details of science, or Darwin's arguments. • Is God the Designer? • Need to prove The Designer(s) = God. • Roman Catholics under Pope JP II accepted the theory of evolution for the Body but not soul • Fundamentalist Protestants reject NS. • Note: Proof that God is not the Designer does not show that God does not exist. Just that he was not active in the way alleged in Genesis, viz., made each species selectively. • Do we know the mind of God? Do we know why or how he made each species? This is a Blank. • Only (allegedly) know THAT God created, not HOW or WHY. No mechanism postulated • Tricky Designer TD Hypothesis • TD: God made the World and species; He also made it appear as if it evolved geologically & biologically - but they did not. • Bertrand Russell Hypothesis - World created 5 minutes ago! • Counter-argument: Prediction and Explanation in the sciences. • Evolution E and Genetics Gen: E+Gen is a predictive and explanatory science • From God hypothesis G no prediction and explanation is possible. G explains nothing! • So E+Gen explains and predicts whereas G does not. G-theory adapts itself to the facts. • Lakatos: degenerating programme of G-theory (he means it is not leading to new discoveries and new facts) • Best explainer: D, TD, R, E? R, D, TD do not explain? • Idea of Sequence of Small Changes. • Darwinian evolution by a sequence of small changes. • Differ over speed of changes. Darwin thought they were slow. But evidence of some very fast changes available (Darwin's finches). • Speed of change is a sub-hypothesis to Evolution by NS. • Species S in two groups S* and S** • S* has advantageous variation and off-spring inherit and survive • S** lacks this and have no future off-spring. • Speciation takes place when S evolves into S* • Lamarck's False Theory • Can characteristics acquired during one's life be passed onto off-spring? • Giraffe stretches its neck to reach higher leaves; passes longer neck onto offspring? • Muscle builder passes bigger muscles to children? • Problem: how do these activities affect the genes that are passed on? They cannot. • So acquired characteristics are not passed on. • Genotype affects Phenotype; Phenotypes do not affect Genotype. • Darwinian Evolution relies on variability in gene pool. • Giraffe Neck and Sexual Selection • Giraffe grazes at all heights; males have longer necks than females. Facts of grazing make Lamarck wrong and Darwin less correct. • Is stretching theory remotely right? • Is sexual selection at work? Not Natural Selection? (sexual selection are characteristics that have to do with reproduction) • Not a closed question, but the common explanation is this: Male giraffes fight by clubbing opponents with their large, massive heads; the neck plays the role of a muscular handle. The largest (longest-necked) males are dominant among other male giraffes and mate more frequently. • Since long-necked males mate more frequently, selection works in favour of long necks. • Nat.S (can be used to mean environmental factors) and Sex.S are rival theories within Darwinian Evolution. Need more evidence in favour of NS or SS. • Idea of Evolution as Small Changes [Dennett on Algorithms (§2.4)] • Algorithm is a process for determining an outcome. • Examples: • Multiplication in arithmetic. • Long Division in arithmetic. • Determine winner of race - who comes in first. • Determine who buys at auction - highest bidder. • Determine who buys at auction - toss coin. • Determine who buys at auction - hair colour! • Tennis Ladder Tournament as Algorithm • Series of games against various teams is an algorithm for determining best team • Not deterministic in the sense that if you run the same process, you don't get the same outcome • Tennis ladder decided by: • Who comes first. • Who comes last. • Who produced most aces (but may not win) • Who was nicest to the umpire • Toss of coin • Gets greatest applause • etc • All of the above are ways, methods, algorithms for getting a result • Evolution by small steps: • NS is Algorithm for deciding steps • Speciation: Old species S evolves into new species S* • Nat.S and Sex.S act on variation in phenotype (and so genotype) so that S becomes S* • Example: drought and beak thickness of finches • Enough small changes, and S evolves into S*. • At what speed? Fast? Slow? • Tennis tournament takes time! • Darwin thought it was slow; but can be fast. • R. Dawkins on Small Changes • Hamlet : methinks it is a weasel • How many generations to reach target? • Target is a phenotype that is advantageous to the environment • Further Issues • How can complex organisms evolve from earlier more simple organisms by small steps? • The eye is complex. Did it evolve by small steps? • Eye = organ receives information by light. • Fact: eye has evolved not once but about 40 times in course or evolution. Eye just suddenly appears • Human eye is a bad design; octopus eye is better! • Problem of complexity evolving out of simplicity. • Evolution or Intelligent Design, or Creation?

• Write your own account of the argument Darwin advances in chapter 13 section 5 on blind cave insects o Facts: F1 = Not strong similarity between blind insects in caves of Europe and America F2 = similarities between insects in caves and those outside in their immediate environments o Probabilities SC = each species results from a separate and special creative act of God F1 is unlikely given SC since there is no reason to suspect the creator would want separate designs and lacking such reasons. Given the similar environment, we would expect that God would want a similar design, but this is not what we observe F2 is not likely given SC (f2,SC<1/2) since perhaps Darwin expects the creator might be expected to create different designs for the internal and external insects given the different environments. o Probabilities NS = Species evolved from a common origin through variations and natural selection F1 is likely given natural selection F2 is likely given natural selection since we would expect similarities if internal insects evolved from external insects o Conclusion NS should be favoured of SC since makes the probability of the facts higher than SC • Do you think it is correct? o Yes I do think it is correct o NS conciliates the facts and does confer a higher likelihood upon the facts then SC o The theist could object that God may want to discard the principle of similar designs for similar environment o But that would appear to give a misleading impression o If God is loving, then we would not expect him to want to give an misleading impression. o Therefore, in the absence of evidence which suggests the contrary, it seems God would have little motive to abandon the principle

• Outline: • Two creationist theses • Objection from imperfect design • Testability • Two responses from creationist • Two "Creationist" Theses 1. Young Earth (YE) Creationism. Bible literalism in which Earth is no more than about 10,000 yeas old • Bishop James Ussher (1581-1656). Ussher Chronology. Earth created on late afternoon Saturday 23 October 4004 BC. Noah's Flood 2349-2348 BC. Etc. 2. Minimal Intelligent Design Thesis (Mini-ID): • There exists an Intelligent Designer, D, Such that for all complex adaptations A (e.g., animal bodies, eyes, brain, circulation of blood system) D created A. • Mini-ID different from YE Creationism • ID-er supernatural but not immediately identified with God. • Objection of Imperfect Design 1) If an omnipotent, omniscient and omnibenevolent (OOO) Creator, Intelligent Designer, [God] exists, and creates organisms, then they have optimal design. 2) Organisms have features that are not optimal. 3) Either: ID-er [God] either did not create these suboptimal organisms; Or ID-er is not omnipotent, omniscient and omnibenevolent. • Creationist might respond that we don't know the creator's wants • Popper's definition of science: falsifiability • Sober has a look at this • Definition: A statement S is testable/scientific = Either (i) S can be directly shown to be true or false or (ii) S can be indirectly examined through its consequences which can be shown to be true or false. S is not testable if S does not satisfy (i) or (ii). • That is, S rules out some claims as being true. S is testable if it is possible for S to conflict with what we observe. • Note: Being testable is not the same as being true or false. S can be testable yet S is false. (Sober misses this?) • Is Mini-ID testable? Is mini-ID testable directly or only testable by means of its consequences? • Testing Mini-ID • How does one test Mini-ID?? • Always need a) a Hypothesis T to be tested and b) an auxiliary assumption A applied to T, viz., (T&A) • E.g. • (i) T = ID-er created all things, & • (ii) A = ID-er wanted them to be purple (A = auxiliary assumption about ID-er wants). • C is Testable Consequence of [T&A]: C = Everything is purple • C is false: there are some non-purple things. • E.g. • (iii) T = ID-er created all vertebrates & • (iv) A = ID-er wants them to have eyes. • Testable - and true! All (normal) vertebrates have eyes. • So ID theory is testable? • Sober p 5: on Duhem Thesis • Omit section on testing probabilistic statements. • Testing is comparative in this context. He is looking at evidence for one hypothesis over another. Sober's idea of comparative likelihoods or IBE. (Some testing not comparative?) • Duhem's Thesis: When Testing any hypothesis T we need to adopt auxiliary assumptions, A. Test (T&A). • Example • Wish to test Newton's laws of motion (T). • As auxiliary A, pick planets in solar system and add in information about their mass and position. • Draw out consequences C and test them against observations. • If (T&A) → C and C fits observation O, then (T&A) pass! If C does not fit with O then (T&A) fail! • So (T&A) together provide consequences which confront the tribunal of evidence/observations. Applies to all theories. • Testing the Auxiliaries? Independent Support. • Cannot test T by itself; need auxiliary A. • But is A testable? A is a hypothesis about ID-er's wants. • A needs to be tested independently of assuming T. Cannot use an untestable, or untested, A to help test T! • A logical point: NO Trivial Auxiliaries! (Sober p6) • Want to test (T&A) and we observe O. • So pick an A to be either not-T or O. • Logically, from (T&[not-T or O]) we trivially infer O! So prohibit this trivial move! • Problem for mini-ID: to find non-trivial, independently testable auxiliaries to test mini-ID. • Problem is that there are none about the auxiliary concerning ID-er's wants!! (Impious to know about ID wants? What independent evidence could we have?) • Lack of independent test of A • Fact F = frigate birds live off the sea yet they lack webbed feet. • Does F support the following claim about Creator's wants and intentions? • A = The Creator wanted there to be frigate birds which live off the sea yet they lack webbed feet. • To so argue is to make the following non sequitur: • F: frigate birds live off the sea yet they lack webbed feet; • A, the Creator wanted that it be the case that F. • Can the inference be improved by adding further premise(s)? • It is hard to see what the extra premises might be. • As well as F, need further premises about: • (1) the existence of a Creator and • (2) the Creator's omnipotence as far as exercise of abilities is concerned. • This begs question about an issue between NS and SC, viz., how distinct species came into existence. • So fact like (F) provide no support for claim A = the Creator wanted (F). Need grounds, independent of the bare fact that F - but none are available. • Some responses of Mini-ID-ers • Mini-ID*: There exists an Intelligent Designer, D, such that at some time t in the history of the evolution of complex adaptation A, ID causally intervenes at t in making of A (but does not intervene otherwise). • Example: In production of newspaper is the typesetter the ID-er? • Mini-ID* is untestable: if no ID-er intervention found as far back as time t in the past causal history of A, then there is always an earlier time t* to investigate (even as far back as the first moment of creation - if there is one). This is untestable. Can never prove this false. Nothing can count against it.

What is the problem of Irreducible complexity and a moral? • Problem of "Irreducible Complexity" • Definition: A system S of interacting parts is irreducibly complex (ICS) = the removal of any one part of S causes S to cease functioning. • Examples of ICS: • mousetrap; arch with capstone (roman arches with scaffolding); • blood clotting system; bacterial flagellum (paddle, rotor and motor) etc. • NS puts emphasis on idea of small changes. • Existence of an ICS falsifies/refutes NS? There are no small changes that can lead to an ICS. • Some alleged cases of ICS have been shown to fit with NS • Final Moral • NS is not without its problems. Exciting to solve them as science progresses! • Is there a proof that there is some problem that NS will not solve? Being unsolved now does not mean unsolvable in principle (or ever unsolvable). • Rivals to NS, SC or mini-ID, have more problems: • In comparison with NS, SC (mini-ID) is a much weaker explanation. • SC and Mini-ID have the problem of finding independent evidence for what a Creator wants that are a central part of their explanation - but there is no independent evidence available.

What is the Lakatos's view? • Two differences between Lakatos and Popper o Popper considers a hypotheses of set of hyptoeshes as scientific or not whereas Lakatos considers a sequence of hypotheses as scientific o Popper says a falsified statement should be rejected from science whereas Lakatos describes conditions under which falsified hypotheses can be kept and be progressive • Background to SRP o Used by Popper but not developed until Lakatos o Uses the idea to try give a definition of science • SRP: o Sequence of theories of whatever size o Each theory has a "hard core" and auxiliaries o Hard core is assumed as true, (conventionalist twist) o If HC is not accepted, it is a sign that it is a new SRP o All theories have the same hard core but different auxiliaries and o They make different predictions (so its theoretically progressive) o Note the change in theory as not just the hard core but also its auxiliaries • Examples of hard cores: o Newton's three laws of motion and the law of gravitation o Copernicus that sun is stationary wrt to the stars • Heuristic o What the scientists consciously or unconsciously do o Heuristics develop the sequence : Negative aspects saves hard core from change Positive aspect governs changes to the auxiliaries in coping with internal problems or external problems from data o Negative heuristic Do not direct criticism/refutation of the SRP at the hard core but at the protective belt of the auxiliary o Positive heuristic Directives - Rules, hints, suggestions - which the scientists follow about how to develop the auxiliaries/theories Changes depending on the context and may involve mathematical equations, rules for developing models, etc. Model is an abstract idealised theory which is complicated with auxiliaries A programme may develop simply by applying the positive heuristic and solving its own internal problems without reference to data Example of a behaviouristic research programme not developing its theories in accordance with the directives of cognitive science • Anomalies o Might be some data that is inconsistent with the theory o Examples: Einstein's relativity theory Newton's model Bohr's theory of the atom All of these made new predictions o Lakatos says it is not necessarily refuting and that theories grow in a permanent ocean of anomalies • Ad hocness o Ad hoc 1 = adds no theoretical or empirical content o Ad hoc 2 = add theoretical but not empirical progress (only add false predictions) o Ad hoc 3 = is any means to solve a problem that arises within an SRP that is not permitted by the positive heuristic of the programme • Application to history o One can engage in the rational reconstruction of any episode of science within the models, but it may be that the scientists themselves hadn't thought of their work in such a way • Appraising the SRP methodology: is it good or bad or accepted or rejected • Progressive or degenerating problemshift o This concerns how the programme deals with its problems and whether its shifts are progressive or degenerating • Theoretical progress o T* is theoreticcaly progressive wrt predecessor T such iff T* makes some predictions T does not make (excess empirical content) • Empirical progression o T* is empirically progressive wrt T iff T* makes some true predictions which T does not make o A T* can be kept and theoretically progressive even if it has been falsified • Empirical Stagnation o T* is empirically stagnating wrt T iff T* is theoretically pregressive but not empirically wrt T • Theoretical stagnation o T* is theoretically stagnating iff wrt T T* has the same empirical content • Theoretical degenerations o Iff there is less predictions • Emperirical degeneration or degenerating problemshift o Iff T* is not empirically or theoretically progressive and T* explains (if it does) any new fvacts by ad hoc adjustments to its predecessor • Rejection (Elimination) and Acceptance o Popper says theories can be rejected from current science when falsified or superseded by more falsifiable theory o Many have commented that science would be denuded of most of its hypotheses if this was strictly followed o Instead Lakatos proposes two types of rejection T is superseded or rejected = there is another T* in the SRP which is empirically progressive wrt. T T is superseded or rejected = none of the novel predictions of T wrt earlier Ts are found to be true o Theories are accepted in an SRP = they are shown to be empirically progressive • Doom of all scientific research programmes o Lakatos thinks they all succeed for a while and then fail o All theories are 'born refuted and die refuted' o "Anomalies, like an odd shaped tree at the seaside, or a serpent seen in a dream, may provide psychological stimulus for progress." Hence, they shouldn't be ignored. • Novel fact o Some true prediction no made prior to T o A fact can be known prior to T but not explained by any prior SRP so therefore the fact is explanatorily or logically novel in that it now has an explanation rather than remaining as an anomaly • Demarcation o An RP is scientific iff has at least one novel fact or is intermittently empirically progressive o Then there are degrees of scientificness • Problem for Lakatos • Case study o Four theories where each avoids the false predictions of the others and makes new predictions, each of which are false o It makes theoretical progress in make new predictions, but no empirical progress o Suppose that, instead, someone goes from t1 to t4 to get three novel facts o How come the first crawl is not scientific but the jumping is. Start and end points are the same o What happens if the scientist does the crawl but only publishes the leap o Lakatos says we must investigate journals, notebooks etc, to make sure there are no leaps and appraise it as scientific o Nola thinks it is an unresolved problem • Problem is that the theories have no empirical progress of novel facts and it is not a scientific research programme • Problem, how come one can lea • Three kinds of programmes - progressive, stagnating or degenerating o Lakatos doesn't give you advice as what to do or whether to go with one type o Just classifies programmes o Lakatos is a narrow methodology

• Sober's account o Oganisms are intricate and well adapted o There are two explanations for this: an intelligent designer or random forces o If we were to come across the functional complexity of a watch and assume it had a designer, then we must prefer at least as much the explanation that the Organisms had a designer o Uses the likelihood principle • Hume's account o Watches are the product of intelligent design o Watches and organisms are similar o Organisms are the product of intelligent design • Personal view of argument o Parts are framed and put together for a purpose o He states it doesn't matter if we have never seen a watch made or known anyone capable of making it o This indicates that he considers sample size irrelevant to inferring design in the watch and therefore he would consider it irrelevant in the case of organisms o We should be charitable and these is no reason, as Sober points out, to consider this an inductive analogy o Instead, it is more charitable to interpret the argument as being an abductive inference

What is the Kuhnian Model for development of sciences? 1. A paradigm PR. is established 2. Normal Science within a paradigm (exemplars and problem solving) 3. Anomalies (Uranus and Mercury for Newton's laws) 4. Crisis (arising from anomalies) 5. Scientific Revolution (new paradigm to cope with the anomalies) • 1* Establish new paradigm PR.* with its Normal Science → 2* → 3* → 4* → 5* • 1** Establish new paradigm PR.** with its Normal Science → 2** → 3** → 4** → 5** • and so on. Phase 1: Paradigm - what is it? • Ancient Greek paradeigma (from verb to show). Example, illustration, a central case as a model for other like cases. • E.g. showing someone a red chair and saying that is the paradigm of red • In English: some central cases to illustrate a general pattern in a wide range of cases. • Kuhn introduces word into philosophy of science in these senses, and many others! • Margaret Masterman: distinguishes 21 different senses, or uses, of word "parafigm" in SSR! (Lakatos & Musgrave) • In Postscript Kuhn admits to 2 senses when he drops the word in favour of Disciplinary Matrix (DM) + Exemplar • Kuhn's term now in common usage in science, history of ideas, etc, - though he explicitly abandoned it later, and never used it in his historical work. • Very unclear concept - though suggestive. Phase 1: Paradigm (i) 1. Identification by concepts - Framework of concepts. PRs identified by system of concepts they employ • Pre-Newtonian Science: concept of heaviness • Newtonian Science: inertial mass (reluctance of a body to be moved) v. gravitational mass (force of attraction). [Q: how does concept heaviness fit in, if at all?] • Einstein relativity: Rest mass v. relativistic mass (when in motion in a frame of reference). [Q: how does Newtonian mass fit in, if at all?] 2. Mutable List - A PR is at least a list of concepts (which can under go change from paradigm to paradigm (e.g., 'mass'). • These Concepts systematically linked in a network to one another to form sentences of a theory, a linguistic network, or a system of beliefs, e.g., laws. • Change in a paradigm is at least a change in the networking of concepts. 3. Lens of perception or interpretation - A way of seeing, or interpreting in the light of the paradigm. A way of organising data, phenomena (experience). Kuhn says observation is "theory-laden' or 'paradigm laden'. No neutral observational language. 4. Rival Paradigms - Perspectivism - each science done from a particular perspective. Science as a host of perspectives. Rival? How so? How radically different? 5. Science as a sequence of changing paradigms. • SSR p 92: 'Scientific Revolutions are non-cumulative developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one.' • How do they change? Why? Why accept one paradigm rather than another? Is there nothing rational about this? Are there paradigm-transcendent values & methods? Phase 2: Normal Science (NS) PR + NS • Paradigms as shared examples - exemplars. • I - Set of exemplars (EX) to illustrate the PR (but do not explicitly characterise the PR.) • From text-book EX as paradigms, we learn to solve other problems in the same way. Text-book exercises as way of entry into the paradigm PR. via EX. 1. S -Exemplars as concrete situations of problem solving. Exemplars are modelled on one another. 2. L - Laws as Exemplars. Newton's Law as an exemplar for other laws: F = md2s/dt2 [Postscript p188] • Free fall: mg =md2s/dt2 • C - Come to see not a wide range of disparate cases with nothing in common, but a large number of cases falling under the same general kind - paradigm. View all cases as of the same dynamical kind. • Examples: • Exemplars in dynamics: Blocks sliding down planes; free-fall; projectiles; motion about a fixed centre, etc. • Simple Pendulum: mgsinθ = -ml d2 θ/dt2 • Harmonic Oscillator; [Postscript p188-9] • Planetary motion about fixed centre; etc. • Do NOT start with Newton's Law of Motion such as: • F = md2s/dt2, i.e., rate of change of momentum expressed as a second order differential! • Get initiation into the Law through exemplars. • Most science = NS conducted within a given PR. • NS as problem-posing and problem-solving within PR. = NS. • NS sets criteria for successful solution of problems (how so?) • Student apprenticeship through exemplars in each distinct science. • Progression of PR: • Is Newtonian mechanics from Newton to Lagrange, Laplace, Hamilton, Mach .. within the same paradigm? • Different growing articulations of what is implicit in PR. Phase 3: Anomalies #1 • Definition: With in a given paradigm PR., AN. is an anomaly with respect to PR. at a given time t =Defn AN. is a problem posed within PR. such that (i) AN is inconsistent with PR., and (ii) no resolution has been found within PR. up to time t. • Need background of PR. against which AN. arises - otherwise not known that a happening is anomalous. • ANs do not immediately refute PR. • AN. might be ignored - only later becomes focus of attention. • Three responses to anomalies. 1. Solution - Some smart person finds a solution within PR. at a time later than t, resolving the anomaly within PR.. • Example: Anomaly of orbit of Uranus with respect to Newtonian Dynamics solved by Leverrier & Adams in mid-1800s with postulation and discovery of new planet Neptune. So inconsistency of anomaly only apparent and is resolved. 2. Relegation - Anomaly AN. wrt PR. is relegated to another paradigm PR.* within which solution of AN. is found, rather than PR.. AN. is no longer anomalous wrt PR.. • Example: Brownian motion discovered in a biological context but later relegated to thermodynamics. 3. Resistance - AN. resists all attempts at a solution with PR. and no new PR.* comes to its rescue. AN. becomes a major issue for PR.. • Example: attempts to solve anomaly of precession of Perihelion of planet Mercury within PR. of Newtonian Mechanics. • Resort to Popperian ad hoc saving moves? • What if different scientists do this in different ways? • They tinker with PR. in different ways - there being no overall agreement about how PR. is to be changed to solve the problem that AN. poses. Phase 4: Crisis • Crisis = There is not just one anomaly confronting PR., but a growing number of them; AN1, AN2, AN3,.. • Division: • There is no agreed way to alter PR. to accommodate each anomaly. • Different Popperian ad hoc solutions for each anomaly adopted by different scientists who previously adhered to the one common PR. • PR fragments, becomes more ad hoc, with different scientists adopting different versions of it. • No unified PR.; so no common NS. Division within community previously united by common PR. • Crisis phase different from NS phase. Scientist no longer agree about the features of their paradigm causing its fragmentation. • Crisis in Newtonian Theory Precession of Perihelion of Mercury. • Observed precession of Mercury of 574 arc-seconds per century. [Complete orbit in 2,250 centuries!] • Leverrier showed (between 1843-59): pull from other planets accounts for only 531a-s/century. Missing 43 a-s/century! What is wrong? Different resolutions of problem • Inter-Mercurial planet. • Inter-Mercurial planet not seen from Earth • Inter-Mercurial stellar dust • Oblateness of the Sun due to its rotation. • Newton's Law of UG not correct (Simon Newcomb). • Not F µ m1m2/d2 but µ m1m2/d2.000,000,1574 • Missing 43 a-s resolved by Einstein's GTR • See papers by N. R. Hanson and George Will Phase 5: Scientific Revolution introduces new PR* • Either: old PR. remains in a state of crisis - any prior unity is a distant memory. • Or: new paradigm PR.* emerges by resolving many or all of the old anomalies AN1, AN2, .. with respect to old PR.. PR* replaces PR. Why? Should it? • We accept new PR* because it solves problems that older PR does not. • New PR* ushers in new exemplars EX*1, EX*2, .. for solving problems posed in PR*. Help to articulate new PR.* - to make new PR* more explicit. • Adoption of PR* a rational process? Or one of mob adoption of PR* and rejection of PR. Elderly holdouts? Is ability to solve problems a value here? What is a Scientific Revolution? • SC = Old PR. is replaced by New PR* • Conceptual revolution: Old concepts employed in PR are replaced by new concepts of PR*. All? Many? Some? Few? concepts replaced. (e.g. concepts weren't replaced) • Law replacement: Old laws ("symbolic generalisations") of PR replaced by new laws of PR.* (Or are old laws modified in some way?) • New laws expressed in terms of new concepts. So conceptual change leads to law change. • Exemplar replacement. They are usually the new solutions to the old anomalies of PR. • If our concepts change, does it follow that what they are about also changes?, So successive paradigms are about different things? Or are concepts like a new "web imposed on the world? Kuhn & Incommensurability • Kuhn & Feyerabend on incommensurability of successive paradigms in an SR • Some numbers such as √2, √3, π, etc cannot be expressed as fractions. They cannot be compared and are incommensurate with respect to one another. Lack a common measure (as a ratio of numbers). • Metaphorically extend idea of incommensurability to a pair of successive paradigms wrt their: concepts; ontology (what the paradigms are about); laws; models; methods; scientific values; observations (!), etc • SRs characterised by radical incommensurability along many dimensions (concepts; methods etc.) • How correct is this wrt actual science? • If two successive paradigms are radically incom., then how can we compare one paradigm with another? Are they about very different things? Incom. observations?

1. Historical - Kuhn criticised for pushing actual history of science into a framework which it does not fit. At a distance OK - but close up it is often wrong. No big changes - One philosopher only saw small changes in the development of electricity theory without saying everything is changing all at once Are there paradigms in the disciplines? 2. Irrationality - What account does Kuhn give of the norms governing what changes we ought to accept in the growth of science? Too much irrational mob psychology? Kuhn's later attempt to talk of common paradigm-transcendent values in Postscript. 3. Circularity problem 1. Defn: Paradigm =Defn what the members of a scientific community share. 2. Defn: a scientific community =Defn set of people who share a paradigm • Circularity! • Kuhn tries to break the circle by taking the idea of a scientific community as basic and then define a paradigm as what they all share. • A sociological approach to identifying scientific communities and then paradigms. 4. A -Kuhn Against Theoretical Truth • Truth at the theoretical level denied (Structure of Scientific Revolutions (extracts in Coursebook #1) p. 206): • Concerning 'the match between the entities with which the theory populates nature and what is "really there" ...There is, I think, no theory-independent way to reconstruct phrases like "really there"; the notion of a match between the ontology of a theory and its "real" counterpart in nature now seems to me illusive in principle. ... no coherent direction of ontological development' • Who do they write to and which conferences do they go to? If it is the same, then they are a same community. Where they did their PhD etc. 5. T - Kuhn and Theory-ladennesss • O is a theory-neutral observational report = for 2 people, P1 & P2, with the same perceptual apparatus and in the same situation for observing but who hold different theories (or beliefs), B1 & B2, P1 and P2 will hold the same observational report O. • O is a theory-laden observational report = for 2 people, P1 & P2, with the same perceptual apparatus and in the same situation for observing but who hold different theories (or beliefs), B1 & B2, P1 will hold observational report O1 but P2 will reject observational report O1 and hold a different report O2 (and vice versa). • Kuhn on Theory-ladenness: What we have learned, or come to believe, through our education and culture, penetrates down to our perceptual apparatus "shaping" what we see, or report we see. • Muller-Lyer Illusion Counter-example: We believe line to be equal; but that belief does not make us see the lines as equal. Cannot do that! • Our visual apparatus resists this penetration by our beliefs • Fodor: Visual system is encapsulated mental module not penetrated by beliefs from outside.

What is the Paradigm. v Discpl. Matrix in Kuhn and science's history? • PS 182. 'Paradigm' dropped in favour of 'Disciplinary Matrix' (DM). [Word 'theory' too narrow.] • DM has 4 elements: • SG: Symbolic Generalisations, or laws (symbolic in a mathematical sense) • M: Models (idealised versions of real systems) • V: Values • Ex.: Exemplars • History of science now a sequence of DMs • DM1 = • Then, DM2 = • Then, DM3 = , etc. • What are the elements of a DM? Do they all change? Symbolic Generalisations (SG): Laws • SGs have dual feature as (1) Empirical claims about how the world operates (2) Definitional claims of what makes them up • Definitional Law • Newton's First Law of Motion in form: F = ma • How are we to understand inertial force 'F'? inertial mass 'm'? Acceleration 'a'? • Velocity v = rate of change of distance over time. • Acceleration a = rate of change of velocity over time. • Can we define 'F' and 'm' by explicit definitions? • No. Only implicitly, or contextually: in the context of the law: F = ma • Thus inertial force and inertial mass can only be understood in the context of a law-like relation. This makes the law non-empirical, or definitional. • Newtonian Mechanics as a conceptual network imposed on the world of phenomena • Other examples of definitional theories: • Empirical discovery or definition? • Water is collection H2O molecules. • Acids have free protons in solution. • Lightning is electric discharge. • Electric current is flow of electrons. • Experiences (having a yellow sense-impression) is nothing but a brain process. • To treat these as definitional is to undermine Popper's anti-ad hoc rule #3. Unlike Kuhn, theories are not like networks imposed on reality that make up the world. • Non-definitional law • Is this true of all laws? I = V/R where I is the current, V is the voltage and R is the resistance. [PS183] • Boyle-Charles Law: PV = RT. (Pressure volume temperature) Can define its terms independently of one another • Improved van de Waals equation: • (P + a/V2)(V-b) = RT. • No implicit Definition of terms within a law. • No change of meaning of terms from Boyle-Charles Law to van de Waals Law. • Kuhn has a Meaning Change Thesis: If the terms of a law are implicitly defined by the laws in which they occur, then the terms change their meaning when they occur in different laws. • Newton's Theory: mass of a body is intrinsic to it and does not alter with velocity of body. • Special Theory of Relativity. Where m is the mass of a body moving with velocity v, m its rest mass. m = m0/√(1 - v2/c2) (v for velocity and c for speed of light. • When at rest the value of m = the value of m0. • But the concepts of relativistic mass, m, and rest mass, m0, are frame-relative. They not the same concept as intrinsic Newtonian mass. • Incommensurability of concepts and referents Models (M) as 2nd element of a DM • Kuhn's talk of 'metaphysical paradigms' becomes models. • Kuhn may well rescind from taking models as pictures of reality or being reality-like. Anti-realism. • Examples • Heat is motion of constituent parts. Heat of a gas is modelled as mean kinetic energy of constituent molecules. • Electric circuit modelled as steady-state hydraulic system. • Different models: (1) Classical Model: perfectly elastic non-attracting very small "billiard balls" • (2) Van de Waals: they do attract one another; also spin, etc. 'Models as a source of analogy. Kuhn's Values as 3rd element of DM • Values - those features we desire/want our scientific laws, theories, and paradigms to exemplify. • All these 9 values provide a common basis for theory choice. • FUCAP PUSS 1. Predictions that are accurate. Prefer quantitative over qualitative predictions. E.g. water is hot versus water is 80°C±5°C, versus water is 81.334°C. 2. Permit puzzle-formation (PF) and puzzle solution (PS) 1. [Compare # of PFs with # of PSs; ratio PF/PS] 2. Puzzle might be how and why do the planets move they way they do and then the puzzle is solved via Newtownian mechanics 3. Simplicity (In Popper and Lakatos) 1. What is simplicity? 1. # entities 2. # Ad hoc hypotheses 4. Self-consistency (internal consistency) 1. Bohr theory of the atom was not internally consistent but he over looked it 5. Plausibility = compatible with other accepted theories (external consistency) 1. Why should this matter if the other theories are wrong 6. Social utility (Would you drop a theory if it had no social usefulness?) • Why adopt these? What is so great about them? • Does Einstein's theory of general relativity have any social utility? 7. Broad Scope: Has consequences that go beyond the observations, laws and theories for which it was initially designed to explained (see Non-Designer evidence). 1. Theories which give you information for which they were not designed 8. Fruitfulness: have novel predictions above those already known. 9. Simplicity as unity: bring together observations, laws etc that would otherwise remain isolated and unrelated (consilience) Values: Paradigm relative or independent? • p. 184-5. More widely shared in scientific community than SGs or Ms. • Give sense of community for all scientists. • Are used in times of crisis to chose between competing viewpoints. Choice between paradigms. • If so - paradigm transcendent and not relative. • SSR p. 42: '... there is another set of commitments without which no man (sic) is a scientist. The scientist must be concerned to understand the world and to extend the precision and scope with which it has been ordered.' • Values of understanding, precision, scope, order 'which have held for scientists at all times' . • So in SSR not all values are paradigm relative but paradigm transcendent. Hold across all paradigms. Some missing values? • Nothing about inductive or evidential support, or Popperian corroboration. • Though scope and fruitfulness might be taken to involve this • Nothing about truth or verisimilitude. • For truth at observational level - "predictive accuracy." But not at theoretical level where it concerns unobservable entities. • PS 206 denies that we can even talk of getting closer to the truth: 'the notion of a match between the ontology of a theory and its "real" counterpart in nature now seems to me illusive in principle". Anti-realism! Doesn't even think one can get closer to the truth with other theories. Same Values - Different weightings • According to Kuhn, scientists can agree on values but weight them differently. • E.g. Prefer accuracy to scope and/or fruitfulness • Prefer simplicity to accuracy; • Prefer plausibility (external consistency) to fruitfulness • Consider 4 values and different weightings for two scientists S1 and S2. Weightings from 0% to 100% • Scientist1 with different weightings for the values: [V1]W1 + [V2]W1 + [V3]W1 + [V4]W1 • S2: [V1]W2 + [V2].W2 + [V3]W2 + [V4]W2 • No guarantee that for any theoretical context, the same overall choices will be made by S1 & S2. • So get incommensurability of theory choice. No "algorithm" for theory choice. • Paradigm shifts just seem to be subjective value weightings Exemplars as 4th Element of DM • Exemplars as concrete problem solutions to certain examples that serve as "paradigms", or exemplars, for solving problems. • Text book model solutions to problems v. take-home exercises for solving like problems. • Laboratory experiments that are standard v. modifications of the experiments in non-standard cases e.g., simple chemical reactions using acids such as HCl extended to other acids. • Laboratory use of measuring instruments such as distance using rulers (as the exemplar), and then by extension theodolites for greater distances, astronomical distances, etc Scientific Revolutions • PR1: • DM1 = • SG1 • M1 • V1 • EX1 • Then • DM2 • SG2 • M2 • V2 • EX2 • What happens if the values remain constant all the way through - make transcendant values across DMs. He later defined science by these transcendant values

Read the paper by Lakatos called 'Introduction: Science and Pseudoscience' in the Coursebok Chapter 19. Describe lakatos' own account of the demarcation of science from pseudoscience and why he thinks it is an improvement over Popper's or any other account: • Lakatos' theory o Scientific inquiry is demarcated from non-science by virtue of whether the inquiry is scientific research programme o A RP involves a series of theories each with a shared hard-core of postulates and differing auxiliary assumptions from which predictions are deduced from the theory and tested o A SRP is one which predicts at least one novel fact during the programmes life time: makes "dramatic, unexpected, stunning predictions" o A novel fact is one which is not predicted by its predecessor theories and is true • Why it is an improvement o Popper Difference - Lakatos differs from Popper in that Popper proposed that once a theory has been falsified, it is then rejected from science whereas Lakatos allows falsified theories to be kept in science under certain conditions Refutations - Lakatos considers this superior since it according with history where theories grow in a permanent 'ocean of anomalies' and Popperian refutations which quickly kill theories are absent Practice - Scientists avoid refutation by either inventing rescue hypotheses to save the theory from refutation or they label the refutation an anomaly and ignore it Examples - The scientific research programmes of Newton and Einstein were refuted instantly but they all predicted novel facts - fact either undreamed of or contradicted by previous or rival programmes o Kuhn K - Kuhn considered scientific revolutions to be akin to a sudden irrational religious conversion and, therefore, there is nothing to distinguish science from non-science -S - Lakatos's theory is superior as history of science refutes the idea of sudden scientific revolutions and it provides a demarcation criterion, D- Demarcating science from non-science has grave social and ethical implications since groups have often made harsh judgments based on a criterion of scientific demarcation • Four laws constituted hard core of the Newtonian programme

What are the epistemic aims of science, a definition of truth, a theory, truth-likeness, and two cases of verisimilitude? Truth as aim of science • Empiricist Aim: In our scientific theories we aim to capture all the true observational claims (and to avoid any false observational claims). • Realist aim for science. This includes the empiricist aim + more: in our scientific theories we also aim to capture all the truths about unobservable reality (and to avoid any such false claims) • Realism supposes that there is a realm of unobservable things and properties which science can tell us about: electrons, DNA molecules, inertia, quasars, etc • While empiricists might be agnostic about an unobservable realm, realists are not. Science can uncover truths about it. Truth 'Definition" or Schema • Aristotle Metaphysics 1011b25-26 • "To say of what is that it is not, or of what is not that it is, is false; while to say of what is that it is, and of what is not that it is not, is true." • What is = What is factually the case. • E.g. Auckland has 2 harbours. • To assert 'Auckland has not got 2 harbours' is to say the false; • To assert 'Auckland has got 2 harbours' is to say the true. General Schema as Definition of Truth: • q is any proposition, eg, 'electrons are negatively charged' • Truth matrix Avoiding Ontological/Epistemic Confusion • Do not confuse: • What there is - ontology - with • what we can know about what there is - epistemology • Example: A cargo boat is passing North head now. • I say [think, suppose, etc]: "a cargo boat is passing N Head now" • Suppose what I say is true. • But I do not know that it is true. • Suppose what I say is false. • But I do not know that it is false. What is a theory? • Nola takes it as Laws of nature: At least generalisations • Examples • All bodies expand on heating (true?) • S µ ½t2 [distance S of free fall µ time squared] for all distances and times. • Laws of reflection: for all incident angle î, there is an angle of reflection ř such that î = ř • Law of refraction: for all î, sinî/sinř = constant. • Theory = two or more laws applied together. • Example: • Newton's three 3 laws of motion and Law of Universal gravitation. • Genetics: Mendel's 2 laws of segregation and independent assortment. • Law of Segregation: during gamete formation each member of the allelic pair separates from the other member to form the genetic constitution of the gamete • Law of independent assortment; during gamete formation the segregation of the alleles of one allelic pair is independent of the segregation of the alleles of another allelic pair Truth-likeness or Verisimilitude • We think that as the sequence of scientific theories grow and develop, that they are getting closer to the truth about how the world is. • Theories of motion: Aristotle à Galileo à Newton à Einstein à; etc. • Analogy for getting closer to the truth • Example: the time is 9 am. But three different people give three different times: 8.45 am; 9.07 am, 11 am. • Note: all three times are false. • Example gives you truth to measure the verisimilitude whereas science doesn't give you what the final truth is • Verisimilitude = measure of closeness to truth of sequence of false theories Theories have infinite consequences • Theories, since they contain general laws, have a large number of consequences, viz., instantiations of these laws. E.g., the instantiations of the Boyle-Charles Law PV=RT • Theories can be applied in many different circumstances. So there is a large number of consequences flowing from each application • H(applied) àP1, àP2, àP3, .... àPn, .... • For convenience let us denote all the true consequences of H (applied?) by 'H(T)', and all the false consequences of H (applied?) by 'H(F)'. Verisimilitude (Ver) of 2 (False) Theories • Blue theory has two true consequences and one false one where one true one is shared with the red theory which contains one falsehood and only one true • Therefore, blue theory has higher verisimilitude • That is Popper's basic idea but it is difficult to make it work Harder Case of Verisimilitude • More difficult case is when two theories have uncommon falsehoods and truths • Popper's naïve idea was that a theory has higher verisimilitude if it has more truths than falsehoods wrt another theory - Popper conceded it was too technical • In previous hard case, the classes of truths and falsities might be incomparable. • There are also some logical problems with the first simple definitions given (as above, initially by Popper).

What is a definition of realism & non-realism? • Realism about some class of entities E: = • (1) Existence Dimension: the Es exist, and • (2) Independence Dimension: Es exist independently of us: that is, if we were not to exist, or we do exist but do not perceive the Es or think or theorise about them or refer to them, the Es would still exist. • REALISM[(1)&(2)] claim: E's exist mind-independently • Example: • Examples of R[(1)&(2)]: ordinary objects, tables, stars, sand, water. Idealists are anti-realists who deny (2) but accept (1). • Properties: green of leaf; liquidity of water; hardness of rock, mass of electron, charge on an electron, etc. R[(1)&(2)]? • Non-Realism: cases of (1) but not (2). • For example: • Psychological: felt experience, pain, tickles; • Social: Money, property, wills, cheques, elections, etc (see John Searle) • Colour realism or anti-realism? Anti-realism about colour - in absence of humans, world is not coloured! No sounds! Mathematical entities - realist about numbers? Two realisms 1. Common Sense Realism: Most observable objects exist independently of us, R[(1)&(2)]. No ghosts, unicorns, etc • Observable objects and their properties: rocks, water, and their properties: 2. Scientific realism = Most postulations of unobservables in science exist independently of us. R[(1)&(2)] 1. Unobservable Objects and their properties: the "theoretical" items of science. • Examples • Electrons, their mass, their charge. Protons, etc. • Inertia and gravitational attraction, gravitational fields. • Atoms and molecules. Tectonic plates. • Quasars, pulsars, galaxies, black holes • Bacteria, viruses, DNA molecules, etc • Some failures: • Phlogiston? Absolute space? Impetus? Electromagnetic ether? Frame-independent mass? Non-realism concerning all of these??

The IBE Argument for Realism 1. Given claim F - viz., T makes some novel predictions P that turn out to be true. 2. TR offers a good explanation of why F is true. 3. TN-R does not offer a good explanation of why F is true. (It is only a model!) It is surprising (a miracle!) 4. So: on IBE, prefer, or accept, TR over TN-R. • Novel facts support the idea that our theory has latched onto the world, or at least some feature of it, and it is this that explains why we have novel predictions that turn out to be true. It would be a miracle if TN-R kept having true predictions! • Duhem affirms likelihoods of 2 and 3. • Example of novel prediction: • Let T be Einstein GTR (General Theory of Realativity) • Let P = the path of light bends as it passes the Sun. • P is a novel prediction for T with respect to prior theory B =Defn • the probability of P given B is low while the probability of P given T is high. • P is novel prediction for T (wrt to B) =Defn • [prob(p, T) - prob(P, B)] is close to 1. • Eddington checks out P - it is a fact. So novel fact P gives high corroboration to T. Alan Musgrave version of IBE & Realism (1) Fact(s) F are observed to be the case; (2) Hypothesis H explains F; (3) H satisfactorily explains F (i.e., meets a minimum threshold for being satisfactory); (4) No available competing hypothesis H* explains F better than H; (5) If any hypothesis is a satisfactory explanation of some facts and is the best available explanation of those facts, then it is reasonable to accept the hypothesis as true; (6) it is reasonable to accept H as true. [deduction]

Instrumentalism as a Non-Realism • Instrumentalism: Theories are merely (mathematical) devices for predicting observational outputs given observational inputs. • Theories as "black boxes" are not taken to be literally about anything; they are rules for calculating outputs from inputs. As rules they have no realist import - cannot even be true or false. • Input observation info ---> Theory as calculating device ---> Observational outputs Bas van Fraassen: Realism & Constructive Empiricism • Scientific realism SR: "Science aims to give us, in its theories, a literally true story of what the world is like; and acceptance of a scientific theory involves the belief that it is true." • Constructive Empiricism CE: "Science aims to give us theories which are empirically adequate; and acceptance of a scientific theory involves as belief only that it is empirically adequate." • A theory is empirically adequate (EA) = what a theory says about the observable things/event/data in this world is true. I.e., a theory is to model all and only actual observables of this world and not just those we know about. It does NOT model the unobservables. • Realism aims to capture all the actual observables of CE; but more than this - unobservable reality CE as an explanation of novel facts?? • Suppose T is empirically adequate, i.e., T captures all actual observable facts of this world (whether we know them or not). • Suppose T makes a novel prediction P; • Then P is amongst the observable facts that make it empirically adequate. • The explanation is of the form: 1. All of T's observational consequences are true ( it is EA) 2. P is a novel prediction of T; 3. So, P is true • This is like the pseudo-explanation: All cats have whiskers; Tibby is a cat; so Tibby has whiskers • Does this really explain why Tibby has whiskers? Comparing the two explanations • Realism TR: That our theory T about the inner workings of reality is true, or partially true in the right aspects, explains why the prediction P from T is correct. • Non-Realism in the form of CE - TN-R: All our theory says is that it is empirically adequate, i.e., it fits all actual observable happenings. So if P is an observable prediction from TN-R then all that follows is that P is amongst all of the observable happenings. But is this explanatory??

Components of theories • Observation Statement: singular report about an observable item and its properties, magnitudes, etc. temperature at a time; rain over an hour, etc • Observational Laws (phenomenological laws): regularities about observables such as laws of reflection, refraction in optics: sin i/sin r = k. (Light bends through some medium) Laws of thermodynamics, Boyle's law: PV = RT; Mendel's laws. Graphs recording global warming over time. • Theoretical Laws: Generalisations involving theoretical entities (which you can't observe but postulate to explain what is going on): Inertial law, F = ma; Gravitational force F = m1xm2/d²; Ohm's Electrical law I=V/R; etc. Among of current in a wire is the voltage over the resistance. Structure of a Theory & Truth • Theory as 3-tired structure o Theoretical laws L o ß o Observational laws OL o ß o Observational reports O o ß Indicates hierarchy but also logical relations Truth • Observational report O is true iff O corresponds to observational facts/reality. • Observational Law OL is true iff OL corresponds to observational regularity. • Theoretical law L is true iff S corresponds to unobservable reality. Duhem - 2 roles for theories • Duhem gives two roles for theories • Role 1: Theories explain Observational Laws • Role 2: Theories are 'an abstract system which summarise and classify logically experimental (observational) laws without explaining them'. • He defines Explanation 1 = 'to strip reality of appearances covering it like a veil in order to see the bare reality itself' • "veil" for sound: vibration of materials transmitted to molecules of air; • "veil" for light = vibration of ether • Hidden reality • Another exaplanations could be Deductive Explanation 2: A theoretical law L explains an observational law OL = • (1) from L we can deduce OL (perhaps under certain circumstances); • (2) OL is true, or approximately true [a reality condition?] Theory as 'economy of thought"(pp.21-3) • Duhem - theories as explanations are too metaphysical. We can have no independent grasp of "reality"; we just have appearances. Stay within realms of experimental method of science. • Theories as "economy of thought". The infinite number of observational reports can be summarised finitely in an Observational Law. Cannot remember the infinite Os but can remember the finite OLs. • You do not remember all the cases of refraction but do remember law: sin i/sin r = k. • Parallel economy of thought for OLs. Consider all the OLs of Optics: reflection, refraction, double refraction, polarization, interference, etc • Economy of thought obtained by replacing OLs by Ls of optics. • Could view 'economy of thought" as minimal simplicity of expression while maximizing information content. Theory as Classificatory (pp. 23-4) • OLs come in a lump. • Ls help classify OLs into different groups or classes. • Examples: • Laws governing colours of spectrum due to a prism classified with spectrum of colours produced by rainbow. Both due to refraction. Law of refraction explains OL of rainbows and prisms • But colours of Newton ring not due to refraction - but interference. So underlying L classify apparently similar phenomena differently owning to different classification of OLs governing the phenomena. • Beauty of construction of structure of Ls which lead to ordered OLs. Theory as classifier: a work of "art". Well-ordered theory • Nola's idea • Combine (1) classificatory power of a theory with (2) its being an economy of thought. • Definition: A theory T is ordered = T classifies, and makes an economy of thought of, all known OLs (an so Os) and leaves none out. • T is "well-ordered" = T offers the best classification and the best economy of thought when compared with its rivals. • "best" = L has smaller number of laws; L gives easier deductions to OLS and Os, etc. [No ad hoc assumptions to get OLs and Os, etc) Natural Classifications 1 (p 25) • Duhem makes this idea • We make classifications of species according to their homologies - similarities and differences in properties and structures. • If our classifications correspond to, or approach, real relations amongst species, then we have a "natural classification". • Analogy with theories: Well-ordered Ls of our theory have power to classify OLs (& so Os) and provide an economy of thought. • Theories never seem to explain OLs in the sense of getting at realities behind observational appearances. • Duhem (pp 26-7): Physicists have a conviction they cannot explain or shake off - that a well-ordered theory approaches a natural classification, and so reflects an ontological order. • Reason: methods of science that lead to well-ordered theories do not enable scientists to make inferences to how theories relate to reality. [Q: would IBE help??] Duhem on theories: - Theories do not give an explanation of appearances that corresponds to reality, but rather satisfactorily explains OLs. - Theories represent and classify Ols - Physical theory "never reveals realities hiding under the sensible appearances"