The empirical-analytical method
- Empirical: scientific research based on systematic observation (observation)
- Analytical: decomposable into logical, elementary statements (proposition)
- Results: hypothesis about empirical regularities (expressed as a ‘law’)
Basic principles of the EA method:
Free of values - Third-person perspective - Focused on objective knowledge - Use of statistical analysis This is the common-sense view of science
Logical Positivism
How can we construct a view of science where all ambiguity is removed > a perfect
scientific language.
➢ Creating the ultimate ‘outsider perspective’
Vienna Circle and Logical Positivism:
AKA:-
- “Wiener Kreis”(Carnap, Neurath, Schlick)
- “Logical empiricism”(Nagel, Hempel, Ayer)
Motivation of LP
= The empirical sciences must replace theological and metaphysical world views –i.e., ‘the
unreasonable powers of church and political ideologies’
Motivating question
What is the relationship between researcher and object of research?
Other important features (assumptions) of LP’s conception of science:
- Classical rationality (- Arguments are only valid if they are the results of: - Logical reasoning, or- Empirical proof)
- Criterion of meaning (Statements have meaning or have no meaning.
- Meaningful statements:
❖ Analytic statements
❖ Synthetic statements can in principle be verified) - Verification and confirmation (Criterion of verification: a theory is scientific if it can be ‘shown to be true’ on the basis
offacts obtained by theory-free observation
➔ Ideal: science is composed of true statements, in particular empirical regularities and
laws.
Criterion of confirmation: a theory is scientific if it can be ‘confirmed’ on the basis
offacts obtained by theory-free observation.
Adjusted ideal: science as certains truth, via inductive logic,to practical certainty.
- Notice emphasis on ‘theory-free observation) - Theory-free observation
➔ Relate to Schutz’s ‘outsider’ perspective (- Observations should be ‘theory-free’- the facts must speak for themselves.
- If not, then perhaps “you see what you want to see”. -‘The facts’ are then biased
Starting point: testing a theory only makes sense if the facts are independent of the theory)
Karl Popper’s philosophy of science
Q: how do we distinguish science from non-science?
Consider:
- Marxist theory of history
- Freudian psychology
- Creationism
These ‘theories’ can not be falsified!
- They are not amenable to empirical testing.
- They can explain-away any and all challenges.
Science is not only theory loaden, but full of values.
The hypothetico-deductive method involves making predictions from singular statements
and universal statements.
The hypothetico-deductive method
Two simple rules;
- The game of science is, in principe without end. He who decides one day that scientific statements do not call for any further test, and that they can be regarded as finally verified, retires from the fame…
- Once a hypothesis has been proposed and tested, and has proved its mettle, it may not be allowed to drop out without ‘good reason’. A ‘good reason’ may be, for instance: replacement of the hypothesis by another which is better testable; or the falisification of one of the consequences of the hypothesis.
Scientific progress is made through the constant falsification of theories
Popper: Against theory confirmation!
Hypotheses can never be confirmed
- The best we can do is provide refutations of theories.
Q: Are two unrefuted theories then on a par?
Q: Can we say one is more likely to be true than the other?
Corroboration: A theory is corroborated if its conjectures resulted in novel predictions that
were not falsified in severe tests.
Review: According to both LP and Popper observation and testing are central to science.
Observation:
- LP: theory-free observation
- Popper: theory-laden observation
Testing:
- LP: confirmation
- Popper: falsification
➔ Both argue that testing scientific knowledge is about facts and not about values.
The ideal of value-free science: 3 features
- Science is autonomous: only scientists (and not politics or business) determine which
research questions are worthwhile. - Science is impartial: moral judgments or ideological views play no role in the
acceptance of scientific knowledge. - Science is neutral: researchers tell how the world works, not how it should work.
Science describes reality and does not prescribe what we should do.
Is science actually value-free?
Example: US Census
Constitutional mandate to count the population every 10 years.
Result determines among others:
- Number of Representatives send to the House of Represent.
- Allocation of federal funding for education.
- Result is of political interest to Republicans and Democrats.
Challenges: Counting people… tricky business!
- Babies are born / people die all the time / people move
- Who is to be counted? Who can be labeled as a ‘citizen’?
- Other methodological problem: how are people counted?- Door-to-door?-Mail-in
questionnaires?
Methodological problem: how are people counted?
- Every method raises the risk of undercounting or overcounting the population (even
when corrected by sampling methods)
- Choose a method of counting →decide whether you think under-or overcounting is
acceptable (political considerations involved!
- Issue: Choice of a method entails/presupposes political value
What kind of values (can) play a role in science? Epistemic and non-epistemic values
- Epistemic values = refer to considerations in the evaluation of hypotheses and
theories, e.g., about precision and scope of theories, models, etc. - Non-epistemic values = refer to moral judgments, ideological views or religious
beliefs part of the grounds that determine the assumption or rejection of scientific
claims.
Constitutive and contextual values
- Constitutive values = are necessary in conducting scientific research. Give shape to
research “from within” in a fundamental way. - Contextual values = form part of the situation in which the research takes place.
Design the research in a ‘random’ way and are therefore not necessary for the
practice of scientific research
Constitutive: Necessary for an activity. Shape the activity from the inside and the
activity cannot go on without commitment to constitutive values.
- Contextual: Part of the environment. May shape the activity, but they are not
necessary to conducting it.
→ Whether non-epistemic values undermine the objectivity of social scientific research
depends on whether they are contextual or constitutive
Three phases in the practice of scientific research:
- The phase before the scientific research starts (phase 1)
- The phase in which the scientific research takes place (phase 2)
- The phase after the scientific research has been completed (phase 3)
Moderate Thesis of Value Freedom
Some set the following condition
Science is objective when:
- Only epistemic values play a constitutive role
- Non-epistemic values (such as moral and political considerations) may only play a
contextual role.
–> Science is objective when only epistemic values are
constitutive of scientific practice; moral and political considerations must always remain
contextual.
Scientific ‘objectivity ’ can be preserved through well-functioning scientific community which possesses (Longino, 1990)
- Recognized avenues for criticism: there are recognized forums for critical
debate (scientific journals, congresses, etc.). - Shared standards: the debate is based on generally shared criteria (e.g.accuracy and consistency).
- Community response: criticism is addressed, positions and assumptions are
adjusted if necessaryin the light of criticism. - Equality of intellectual authority: all members of the community are
considered to be intellectually equivalent.
Samenvatting boek
Objectivity in the sense of freedom from pre-existing value commitments seems necessary for sound social policy making
Social sciences value-laden or value-free > both practical and conceptual dimensions
- Practically: Value-laden research would presumably undermine the usefulness of
social scientific results for social policy purposes. Value-freedom means that scientific
results cannot be contested by those with different political interests. - Conceptually: the issue of value-freedom is about the character of science itself.
Cases where political and moral considerations figure in a scientific dispute, 2 questions:
1. How are the values influencing the science?
2. What values are involved?
With this understanding we’d be able to find appropriate ideal of value freedom
United States has a constitutional mandate to count the population, cause the census
determines the Representatives
- Tricky business for many reasons (e.g., people do not simply line up to be counted)
- May undercount or overcount
- Subject to bias and random error and not free from politics
o E.g. Republicans prefer higher counts in republican states
→ Choice of census method requires deciding what kinds of errors are acceptable
- Clear motivations for seeking value-free social science
- Value freedom requires that values must be completely eliminated
Strong thesis of Value Freedom: Science is objective in so far as values play no role in scientific
research
Dimensions of Value Freedom
→ USA example: Political parties were influencing methodology
General problems with methodology:
- “Type l” and Type ll” errors, or “false positive” and “false negative” results
o Cannot simultaneously reduce both
o Determine best method, which risk is more acceptable
Different methodologies → choosing among different ways to achieve best science
- Epistemic value: contributes to good science, good scientific reasoning, contribute to
objectivity
- VS. → Moral/Political values
Value freedom: excluding all values from science
Philosophers have focused on two potential constitutive roles for values in scientific
practice.
- Moral and political values might influence the justification of theories or the
confirmation of hypotheses - Moral and political values might appear in scientific practice as part of the content of
a theory.
Risk and Error
Hypotheses are never proven definitively by any kind of test; they are only more or less
probable.
→ To decide whether to accept a hypothesis, the social scientist will have to choose a level
for the p-value
→ If we really want to get at the truth, we should always demand maximum probability
before accepting a hypothesis
Three different senses of objectivity:
- Objectivity as freedom from bias
E.g., Advertising claim is not objective - Objectivity as intersubjectivity
Objectivity is contrasted with subjectivity - Objectivity as reliability
Derives from the reliability of scientific methods, results that are likely to be true
Value-Neutrality and Emancipatory Research
Value-neutrality is the thesis that social scientific theories should describe facts, not make
policy recommendations
→ Value-neutrality forbids scientific theories from including statements about what ought to
be done or not done
Many programs of emancipatory social science begin with a critique of ideology.
→ “Ideology” here is understood as the relationship between knowledge, oppression, and
systems of power and authority.
→ In human societies, power and authority are unevenly distributed.
→ Proper understanding of the system of power and oppression requires recognizing its
injustices.
Value Presuppositions and Implicatures
Fact-Value distinction: make clear the distinction between descriptions and evaluations is to
say that evaluative statements include explicitly evaluative predicates like “ought,” “good,”
and their cognates
If value-neutrality is abandoned and values become part of the content of social scientific
theories, then it may seem as if objectivity has just flown out of the window.
Value-laden theories have observable consequences and therefore they can be tested in
ordinary ways.
