Looks like no one added any tags here yet for you.
Three types of cultures
Natural science
Humanities
Social science
Natural science
- oldest intellectual pursuit
- quantitative based
- empiricism - all knowledge is ultimately derived from observation
- interested in regularities
- individual events are applied to classes of events
- physical universe is uniform
→comprehensive, powerful knowledge
- mathematisation, abstraction, idealization
Humanities
- “arts” were ways of doing things without thinking
- became a science ca. 15th century (Humanism, Renaissance)
- studied world:
historical human actors
actions that carry intention+ meaning
texts, artworks, artifacts with meaning
- historical particularities - each event is unique
- distrust generalization and idealization
- interpretation, empathy (reconstruction), hermeneutics (text interpretation)
Social Science
- studied world:
human actors and institutions
behavior
rationality
ritual, cultures
- uses both concepts from humanities and natural science
Nomothetic approach
identifies/describes regularities
formulate generalizations and laws
explain observations through generalizations and laws
typical for natural sciences
seeks causes and formulates explanations
Idiographical approach
understanding the meaning of contextual, unique and often subjective results
mistrusts concepts of cause and explanation
prefers interpretation and understanding
Laws of nature
scientific theories
mathematical models + equations
relations between physical quantities
Laws as paradigm knowledge
seen as highest degree of scientific knowledge
also outside of science
Why Natural Science dominates philosophy of science
few terms that explain more phenomena → simplicity
“simpler” images of world
most developed
iconic role in society
social & historical power
delegates reflection to others
Commonsense view of Science
science is based on facts
facts are claims about the world that can be established through careful use of senses
reasoning takes us from factual basis to laws and theories
the resulting knowledge is securely established and objective
Two scientific activities
doing observations
formulating theories
Relation theory and observation
theories explain and predict observations
observations test theories and help decide between theories
Commonsense (Naive view) view assumptions
Facts are directly given to careful,, unprejudiced observers via senses
Facts are prior to and independent of theory
Facts constitute a firm and reliable foundation for scientific knowledge
Problem of commonsense view 1)
-observations as subjective, passive, fallible
- against the common sense view: what you see is not the same as what I see
→ it depends on knowledge and experience
→ observation statements may differ
→ facts are not unproblematically + directly given to observers
Observations are fallible
- scientists disagree about observations
- background theory & technological advances needed
- sometimes observations are fallible because of theory or technology
→ theories are subject to revision
Problem of commonsense view 2)
Theory-laden observations
- facts do not precede theory
- our experiences often depend on theories we already hold
→ we don’t know which facts to look at if we don’t have a theory (we do observations that help answer our theory)
practical interventions
- make observations more objective
arranging the observable situation in such a way that the observation statement does not rely on subjective/cultural/ perspective influences
a good observation
(active + public but still fallible)
consistency (do it same way every time)
repeatability (someone else can do it too)
compatibility with a good theory
Experiment
practical interventions that isolate the process under investigation by eliminating other influences
good experiments
- compatible with a good theory
- routine, objective procedures
- don’t rely on fine subjective interpretation
- consistent and repeatable outcomes
problems with experiments
- eliminating spurious influences is difficult (need to know a lot about them and how to eliminate)
- can be faulty if knowledge informing them is faulty
Experiments are fallible when
- outmoted by new technology
- rejected because of advancing understanding which shows experimental setup is inadequate
- irrelevant because of advances in theory
Experiments are rejected/inadequate/irrelevant when
- setup does not succeed in isolating process under investigation
- measurement methods used that are insensitive/unreliable
- experiment becomes understood to be unable to solve the question
- theoretical advances: question becomes discredited
How does Science proceed from particular observations to general theories?
Observation → Facts → Theory through induction
Deductive reasoning
→ the logically derivation pf a conclusion from premises
logically valid argument
doesn’t add to our knowledge
statement about all to statements about some
Logical validity
an argument is logically valid if and only if it is impossible that the premises are true and the conclusion is false (if premises are true then conclusion must be true)
Inductive reasoning
(common sense view)
not logically valid
statements about some to statements about all
Underdetermination
when two theories are empirically equivalent meaning that both fit the data equally well → data isn’t rich enough to help us decide between two theories
Solution for underdetermination
- make new predictions → explore where theories aren’t empirically equivalent and then make observations for those that aren’t in common
- pragmatic criteria: a theory might be better than another for reasons outside empirical adequacy e.g. it’s simpler but explains facts equally well
laws characteristics
- mathematical equations
- concise and simple often elegant
- universal in scope
regularity view of laws
→ laws are descriptions that just say what happens to be the case
- true universal generalization about specific events
necessity view of laws
→ laws say not just what the case is but also what must be the case
- descriptions of necessary relations between entities, properties or events
Explanation
→ an answer to a why question
- gives us understanding of why things are as they are
- must be true to have status of explanation (otherwise its a pseudoexplanation)
Explanandum
(explananda)
that which is to be explained
Explanas
(explanantes)
that which does the explaining
Prediction vs. explanation
both can take a deductive form
P1: If A then B
P2: A
C: Therefore B
→ premises need to be true for an explanation but not for prediction
Hempel’s models of explanation
Deductive nomological model (DN)
Inductive statistical model (IS)
DN Model
an explanation is:
- a valid deductive argument
- (formed) from true premises
- includes at least one law or true generalization and description of some particular facts
- provides description of the fact that is to be explained
Structure:
- L1…Ln (laws)
- C1… Cn (Facts)
- E: (Explanandum)
DN model objections
the DN model does not rule out explaining a cause on the basis of its effec
also does not rule out an event on the basis of irrelevant info
→ model is too lax (not sufficiently strict)
IS Model
an explanation is an argument that establishes that the explanandum had high probability of occurring
- uses probable, not certain reasoning (inductive)
- gives understanding for explanandum
IS model problem
too restrictive
→ some good explanations do not make the explanandum highly likely
Lessons learned from Hempel
an explanation should track causes, not merely state sufficient conditions for occurrence (DN model fails requirement)
Causal mechanical model
explanation of event E is a description of part of the causal interactions and processes that led up to E
a causal-historical account
causal interaction
spatio temporal intersection between two causal processes that modifies both = when two objects intersect in spacetime
Causal process
physical process able to transmit a mark in a continuous way - something that is extended in space time
causal mechanical model problem
difficult to obtain a full causal mechanical explanation (too many contributing factors)
to falsify
refute, empirically prove that a hypothesis is false
falsifiability
the receptiveness of a theory to being falsified
falsification
act of falsifying a theory F
Falsificationism
Popper’s claim that a scientific method consists in falsifying a Hypothesis
How to test a hypothesis (Popper)
- usually not testable in isolation (too abstract, theoretical)
→ we have to generate observational implications or predictions which we then test
Hypothesis H implies prediction O
we check whether O is true
we draw conclusions about adequacy of H
Two outcomes of testing hypothesis
O is true
→ but H could still be false (possibly another mechanism caused O)
→ does not give us any guarantee of truth value of H
→ confirmation is not deductively valid
O is false
→ then H must be untrue
→ if mechanism posited by H exists then O must obtain
→ guarantees H is false
→ falsification is a valid argument scheme
Science according to Popper
rejects induction by rejecting confirmation
→ took David Hume’s induction problem to be unsolvable
scientists can only attempt to falsify Hypothesis
Scientific knowledge (Popper)
a series of not yet falsified hypotheses (not a collection of true/confirmed statements)
Scientific progress (Popper)
the elimination of false theories
Where to hypotheses come from
any source of inspiration (dreams, observations, esoteric theories)
two stages of scientific work
Context of discovery
- stage of proposing a hypothesis
- no rules or standards
- de facto thinking process
Context of justification
- stage in which H is tested
- logic and rules
- makes science objective
- de jure defence of correctness of thought
Demarcation of science
(what distinguishes science from pseudoscience?)
practitioners must be able to say which observation would falsify their Hypothesis or which outcomes are excluded by theory → otherwise the discipline is pseudoscience
statements that take risk of being falsified = good and scientific
(theories don’t have to be falsified they just have to be falsifiable!)
Popper criticism
Not as straightforwards as P assumed → H only generates predictions when combined with auxiliary assumptions + it’s difficult to pin blame for failed prediction on a single hypothesis
Doesn’t accord with scientific practice → researchers make ad hoc adaptations to theories in order to avoid falsification + dogmatism can have methodological values → sticking with a theory and making modifications
Popper ignored possible ways to “save” induction and the confirmation of theories → pragmatic justification of induction → corrobaration (acknowledged a week form of confirmation)
Some valuable scientific H. don’t seem falsifiable → may eliminate examples of good science
Pros and Cons of falsification
Pro:
simple,logical model of science
scientists as creative , undogmatic , risk-taking (appealing)
Con:
narrow view of science
underestimates complexity of o and h testing
ignores mechanisms of h confirmation and empirical confirmation
Image of Science before Kuhn
Structure:
- theoretical terms have clear and stable definitions
- empirical data provides objective test of adequacy of theory (Naive view of science, Poppers view)
History:
- growth of knowledge is continuous and accumulates
- all scientists in history share same norms of rationality (norms of what counts as evidence, good observation)
Paradigm
a conceptual framework which shapes thinking and work of scientists and defines a period of “normal science” in a branch of science
Consist of :
assumptions about the world as studied by that science
examples of how to solve problems
a style of theorizing
Life within a paradigm
- scientists solve puzzles within a paradigm by imitating examples
- a paradigm gives clear norms for progress (new discoveries), professional stability/career, coordination, concentration of effort
- strict boundaries to creativity
How paradigms end
- when scientists find radically new data that cannot be explained within the current paradigm
- sequential phases : new data as anomaly; explanation is necessary; current paradigm is inadequate → scientific revolution
Scientific revolution
- scientists focus on new data and sketch a new paradigm based on it
- period of revolutionary crisis (split scientific community in conservative and progressive)
- social process of paradigm shift
Incommensurability
lack of shared standards
not comparable
- according to Kuhn, subsequent paradigms are incommensurable + not just different
3 types:
semantical
observational
methodological
semantical Incommensurability
meaning of terms are gotten from the paradigm and differ depending on the paradigm (different meanings)
→ impossible to translate between paradigms, failure to communicate, no clear logical relations between paradigms
observational Incommensurability
concerning sensory perceptions
→shaped by paradigm we do science in → we see different things in different paradigms
(Gestalt switch, ambiguous figures)
- ontological consequences → scientists exist in different worlds, rejection of realism, “different worlds”
methodologial Incommensurability
concerning norms of rationality and progress
→ each paradigm has their own view
→ no paradigm independent criteria for theory choice or norms for scientific progress