Lecture 5 - behaviouralism

different theoretical approaches have a different meaning of theory behaviouralists would subscribe to a definition like this: theory as orientation about a methodological orientation that tells us how to go about social science

• fits with positivist epistemology, but it is not the same logical positivism: school of philosophy of science (epistemological school that was influential on positivism as an epistemology and thus also on behaviouralism) !not all behaviouralists subscribe to all the ideas that logical positivists have

methodological prescriptions on how we should do political science research it is devoid of substance people behave in patterned ways, it remains to be seen in observation in what way - so methodological, not substantive

• political science should (as far as possible) be done in the same way as natural science: causal claims, isolating variables • movement away from normative theory and description • movement away from the (legalistic) study of formal political institutions (old institutionalism) • movement towards observing individuals/agents and explaining their observable political behaviour • claims to be unbiased, objective, value-neutral approach

why do agents (e.g. people, institutions, parties) behave the way they do? • why do people vote? • why do they vote for party A instead of party B • why do they decide to join a party? • why do some of them give to charity? • why do some shun strangers and immigrants? behaviouralists try to find causal answers

it has patterns we can make predictions based on (complicated) models

No, it is a general perspective used by many disciplines in the (social) sciences any social science that tries to explain human behaviour

Why do people behave the way they do? • Why = focus on explanation/erklaren • People = focus on individuals or collectives (still: collectives seen/studied as groups of individuals) • Behaviors = empirical, objectively obervable

• foundationalist ontology (the world IS some particular way, has some objective features) • positivist epistemology (we can have objective/scientific knowledge) • privileges quantitative methodology

mid 20th century associated with the ‘behavioural revolution’: behavioural sciences movement in the social science associated with the rise of comparative politics (is a type of behaviouralist research) subfield: political behaviour before, they were studying: - formal institutions - political philosophy

logical positivism You have to specify what type of claim you are making three types of analytical statements: 1. tautologies (definitional or conceptual statements: not testable) 2. empirical statements: testable against observation 3. other statements/garbage statements (e.g. normative statements) only type 1 and 2 are scientific (so, they find normative theory as not scientific) empirical statement: we need to know what observation would disprove the claim

• chairs are objects you can sit on + they have 4 legs (tautology) • chairs have 4 legs (empirical statements) • all chairs are beautiful (other statement: devoid of analytical meaning)

empirical theory = is normative theory = ought explanatory = why descriptive = what

it doesn’t define what type of statements it makes: it isn’t testable lots of positivists aren’t as radical as logical positivism: normative statements aren’t garbage, they aren’t scientific

• internally consistent (logically possible) • externally consistent with other accepted theories (as far as possible) • consistent with direct observation (factual proof)

verification = how would we know if it were correct? - induction: they check out with what we expect falsification = how would we know if it were incorrect? - we need to have a clear idea of what would definitively disprove a claim

Popper: we can’t affirm general claims inductively, we must seek to generate falsifiable hypotheses Lakatos (1791): not all assumptions/propositions in theory need to be falsifiable (some core assumptions can’t be) *e.g. rational choice models assume that people maximize utility, but this is not measurable/observable. all types of behaviour might be considered to maximize utility - Popper would reject self-interest: not measurable - Lakatos could be fine with it: it results in behaviour and that is measurable/observable

tautology claim: all swans are white = fine, that’s your definition, black ‘‘swan’’ is not a swan by definition empirical claim: all swans are white = we need to say: if we see a swan like bird, that is black, then the claim is falsified garbage claim: all swans are beautiful = not falsifiable; wat defines beautiful, there is no specific observation that can disprove the claim if it isn’t falsifiable, it is not science, according to Popper

behaviouralism has no main ideas/theories, so you can’t derive political hypotheses from that they get hypotheses from: • other fields/disciplines/research (sociology, psychology) • other theoretical perspectives (e.g. rational choice, normative theory)

what can explain voter choice? behaviouralism looks at other theories to generate hypotheses, it tests it in a behavioural way (e.g. survey research • rational choice theory: voting theory assumes that votes are determined by social economic status • cognitive political psychology explains voting behaviour as the result of framing and imaging puzzle: why do some people remain loyal to a party, while others follow electoral trends? e.g. populism

more probablistic than that it is deterministic leads to problems with Popper: if you say something is more likely than something else, than the theory can’t be disproven with one observation, you’d need to see the exact opposite (-> studying many cases) -> complicated falsification

• brought the individual (elite or mass) and groups into political science • brought behaviour into political science • brought focus on explanation into political science • sought to better the scientific standard of political science research • concern with replicability (allows for verification and falsification)

• mindless empiricism, easily measurable? (is it theoretically important, are there variables and areas ignored, is there a focus on statistical correctness instead of theoretically sound explanation) • objectivity of researcher? • independence of research from researcher? • methodological individualism: focus to much on agency, not enough on structure?

alternative explanations also need to be taken into account this can’t be done experimentally -> statistical measures e.g. voters choosing Trump over Clinton because of party identification, but it might also be because of education level, authoritarian personalities etc.