CH 1: the received view of science
Received/standard view: stage for many developments that were to occur over the course of the
following half-century. Derived from program of logical positivism in philosophy of science: broad
philosophical movement. Logical positivists dominated thinking about philosophy of science. Indeed,
much of current direction in philosophy of science is a reaction against views of logical positivists
Logicism: all scientific language, including mathematics, is an extension of logic
Positivism: meant empiricism -> the idea that knowledge arises out of sense experience
Analytic and synthetic a posteriori propositions
Main aim logical positivists program: demarcare scientific knowledge, distinguish science from
pseudo-science, and to remove metaphysical or imagined content from scientific knowledge.
Demarcation criterion rule: accept only analytic and synthetic a posteriori propositions or statements
as scienfitic knowledge.
Analytic propositions: are tautological (true by definition, ex: “all bachelors are unmarried”, also valid
mathematical and logical propositions are analytic, ex: “1+1=2” and “A->A” (-> means implies)
All other, non-analytic, propositions are called synthetic.
Synthetic a posteriori propositions: if these propositions are shown to be true by empirical research,
ex: “my neighbors dog is aggressive” and “the color of coffee I am drinking is brown” -> true in light of
our experience of real world
Synthetic a priori propositions: introduced by Kant. Third category or propositions whose truth is not
shown by empirical research and that are not true by definition (ex: Newton’s laws, sun of triangles is
180 degrees Euclidean) <= universally true: 1. This assertion challenged by developments maths &
physics, 2. Shown that non-euclidean geometries are mathematically possible & contrary to Newton,
Einstein’s relativity theory assumed a curved physical space (these break throughs crucial events for
logical positivists & philosophers)
As a result of scientific developments: logical positivists denied existence of synthetic a priori
propositions in science, asserted that: all propositions that are not true by definition should be
subjected to investigation by empirical research. Intention in doing so: purify science of
“metaphysical” or philosophical claims about the world that were neither analytic or synthetic a
posteriori.
Empiricism: experience is the only source of knowledge (observation?)
For logical positivists, empiricism consisted of two related theses:
1. all evidence bearing on synthetic statements derived from sense perception, in contrast to analytic
statements, which are true by definition
2. predicates: meaningful only if possible to tell by means of sense perception whether something
belongs to their extension (=must be empirically verifiable)
- predicates include: expressions (“is red”, “went home”)
- predicate in grammer: part of a sentence or clause stating something about the subject (“went
home” in “john went home”, or “is red” in “this tomato is red”. Extension of predicate is set of all
those things of which predicate is trye (e.g. set of all red things). Result: “this tomato is red” is
meaningfull because by looking at it, one can see that this proposition is true.
Synthetic statement is meaningful if it can be judged to be true or false by sense perception (other
words: when it’s empirically verifiable, non-analytic statement meaningful if it’s empirically verifiable)
Verifiability principle: criterion for meaningfulness (meaningless: “god created the world in 6 days”)
Logical positivists regared scientific knowledge as possible only in so far as sensory experiences are
systematically related to one another. They drew clear distinction between syntactics and semantics
Syntactics: deal w/formal relations between signs or expressions in abstraction from their signification
& interpretation (ex: well-formed statements, proof, or consistency)
Semantics: deal w/signification and interpretation of signs or expressions (truth and meaning)
Syntactics and semantics important to logical positivists (saw formal logic)
Logical positivists view of task of philosophy of science: clean up conceptual messes inherited from
past experience by pointing out what were/were not meaningful propositions in properly formulated
empirical science.
Summary aims of logical positivists:
, 1. to formulate precisely central philosophical notions as criterion of meaningfulness (verifiability
principle) and distinction between analytic claims (that are true by definition) and synthetic claims
(that must be testable)
2. to develop precise definitions of central scientific notions such as theory, explanation,
confirmation, etc.
Theory and evidence
Logical positivists made distinction between context of discovery and context of justification
Context of discovery: distinction should be drawn between the way in which a theory is discovered
Context of justification: rational reconstruction of theory according to tenets of logical positivism for
the purpose of its justification
Logical positivists argued that philosophy of science should only concern itself with the context of
justification. Context of discovery irrelevant to establishing scientific value of a theory (means: no
relevance to meaningfulness)
Another fundamental distinction by logical positivists: between theories & evidence, facts and data,
since theories depend upon latter for their justification
Scientific theories: systematic collections of concepts, principles, and explanations that organize our
empirical knowledge of the world. In advance of scientific knowledge: theory & evidence given
different weights and play different roles => main problem for philosophy of science & also economic
methodology based on logical positivists thinking, is to explain the relation between them.
Syntactic view: according to logical positivists understanding of theory, proper characterization of a
scientific theory consists of axiomatization in first-order formal language
First-order formal language consists only of symbols representing variables (x,y,x…), function symbols
(A(.), B(.), C(.),..), predicate symbols (A,B,C,…) and the – symbols (“not), v (“or”), A (“and”), ->
(“if…then”), A(upside down) “for all individuals, E (opposite) “for some individuals”
Axiomatization reduces a theory to a set of axioms in the manner of Euclid’s treatment of geometry.
Axioms are formulations of logical laws that specify relationships between theoretical terms. These
terms are implicitly defined by syntactics (=logical structure of theory). As result, language of theory
divided in two parts: (i) observation statements: (or “protocol sentence” that describe observable
objects or processes) (ii) theoretical statements
Meaning theoretical terms given by their observational consequences. Theoretical terms for which no
corresponding observational terms are regarded as meaningless
Theoretical terms identified w/their observational counterparts by means of correspondence rules
(=rules that specify admissible experimental procedures for applying theories to phenomena)
Operationalized: when theoretical terms defined completely in observational terms. According to
received view, for proposition to be meaningful necessary that all theoretical terms are
operationalized.
Example operationalization: “unemploymeny,” which is number of people who are jobless, looking for
jobs, and available for work. -> in this case easy to say who is jobless. However, definition of terms
“looking for jobs” and “available for work” involves human motivation -> cannot be observed as
physical fact, such as oranges being orange. Means, in order to measure unemploymeny, “looking for
jobs” and “available for work” need to be operationalized. This is done by defining “unemployed
persons” as:
<- according to this process,only at this stage of operationalization
can unemployment me measured.
Related to this idea of operationalization and emphasizing idea of meaningfulness, bridgeman
developed extreme form of empiricism (called operationalism). True meaning of a term is to be found
by observing what a scientist does with it, not by what a theory says it does or what its author thinks
, it does. Ways a term is measured. Consequence of this view: term will have different meaning when
it’s measured in a different way:
The nature of scientific explanation
explanation answer to a why question
received view of scientific explanation is more specific: should show some event or regularity to be an
instance of a fundamental law
Hempel developed this view systematically into deductive-nomological (DN) model of explanation (or
covering-law model)
In DN explanation, statement of what is to be explained (explanandum) is deducted from a set of true
statements that includes at least one law (nomos). This latter set of statements called the explanans:
- solid line represents deductive inference
- model explanation of, for ex: why firm x raised its price if one interprets symbols as follows:
- problem w/Hempel’s definition: according to this model, not all deductive inferences can be considered to be
explanations. Consider following ex:
- if assumed George is man, nobody would regard argument above as explaining why George does not get
pregnant. Doesn’t matter whether of not George took birth control pills
- in this instance taking pills not causally relevant for explanation of George not getting pregnant
- explaining law should identify causally relevant factors, but Hempel’s DN model, which focuses on (logical)
deduction from a law, does not require this.
There is an even more fundamental problem w/this model of explanation: required that
generalizations in the explanans be laws. How can we be certain this is the case?
-> closer consideration what laws are: generally suppose law is true but, for current discussion this is
not relevant -> refer to “law-like” statements as statements that are just like laws (don’t know if true).
Idea of “law-like” is just idea of regularity. Problem=many regularities appear to be accidental.
As a 1st attempt to distinguish law-like statements from accidental regularities, see whether a
syntactic definition of law-like statement might help. All law-like statements, then, have logical form
of universal condition:
Difference between two examples: accidental generalization contains designations for a particular
individual object and for a definite date or temporal period, while universal law does not. So
necessary requirement for universal law is that the domain of the antecedent is not restricted to
things falling into fixed spatial region or particular period of time. Law-like statements are required to
be unrestricted generalizations.
Another difference between two examples: implication arrow (->) in syntactic expression of universal
law above denotes a necessity. All elements in antecedent must necessarily also appear in
consequent. This is not the case for accidental generalization: a coin put in smith’s pocket during time
period a would not necessarily turn into a dime -> so law0like statement has no exception.
Received/standard view: stage for many developments that were to occur over the course of the
following half-century. Derived from program of logical positivism in philosophy of science: broad
philosophical movement. Logical positivists dominated thinking about philosophy of science. Indeed,
much of current direction in philosophy of science is a reaction against views of logical positivists
Logicism: all scientific language, including mathematics, is an extension of logic
Positivism: meant empiricism -> the idea that knowledge arises out of sense experience
Analytic and synthetic a posteriori propositions
Main aim logical positivists program: demarcare scientific knowledge, distinguish science from
pseudo-science, and to remove metaphysical or imagined content from scientific knowledge.
Demarcation criterion rule: accept only analytic and synthetic a posteriori propositions or statements
as scienfitic knowledge.
Analytic propositions: are tautological (true by definition, ex: “all bachelors are unmarried”, also valid
mathematical and logical propositions are analytic, ex: “1+1=2” and “A->A” (-> means implies)
All other, non-analytic, propositions are called synthetic.
Synthetic a posteriori propositions: if these propositions are shown to be true by empirical research,
ex: “my neighbors dog is aggressive” and “the color of coffee I am drinking is brown” -> true in light of
our experience of real world
Synthetic a priori propositions: introduced by Kant. Third category or propositions whose truth is not
shown by empirical research and that are not true by definition (ex: Newton’s laws, sun of triangles is
180 degrees Euclidean) <= universally true: 1. This assertion challenged by developments maths &
physics, 2. Shown that non-euclidean geometries are mathematically possible & contrary to Newton,
Einstein’s relativity theory assumed a curved physical space (these break throughs crucial events for
logical positivists & philosophers)
As a result of scientific developments: logical positivists denied existence of synthetic a priori
propositions in science, asserted that: all propositions that are not true by definition should be
subjected to investigation by empirical research. Intention in doing so: purify science of
“metaphysical” or philosophical claims about the world that were neither analytic or synthetic a
posteriori.
Empiricism: experience is the only source of knowledge (observation?)
For logical positivists, empiricism consisted of two related theses:
1. all evidence bearing on synthetic statements derived from sense perception, in contrast to analytic
statements, which are true by definition
2. predicates: meaningful only if possible to tell by means of sense perception whether something
belongs to their extension (=must be empirically verifiable)
- predicates include: expressions (“is red”, “went home”)
- predicate in grammer: part of a sentence or clause stating something about the subject (“went
home” in “john went home”, or “is red” in “this tomato is red”. Extension of predicate is set of all
those things of which predicate is trye (e.g. set of all red things). Result: “this tomato is red” is
meaningfull because by looking at it, one can see that this proposition is true.
Synthetic statement is meaningful if it can be judged to be true or false by sense perception (other
words: when it’s empirically verifiable, non-analytic statement meaningful if it’s empirically verifiable)
Verifiability principle: criterion for meaningfulness (meaningless: “god created the world in 6 days”)
Logical positivists regared scientific knowledge as possible only in so far as sensory experiences are
systematically related to one another. They drew clear distinction between syntactics and semantics
Syntactics: deal w/formal relations between signs or expressions in abstraction from their signification
& interpretation (ex: well-formed statements, proof, or consistency)
Semantics: deal w/signification and interpretation of signs or expressions (truth and meaning)
Syntactics and semantics important to logical positivists (saw formal logic)
Logical positivists view of task of philosophy of science: clean up conceptual messes inherited from
past experience by pointing out what were/were not meaningful propositions in properly formulated
empirical science.
Summary aims of logical positivists:
, 1. to formulate precisely central philosophical notions as criterion of meaningfulness (verifiability
principle) and distinction between analytic claims (that are true by definition) and synthetic claims
(that must be testable)
2. to develop precise definitions of central scientific notions such as theory, explanation,
confirmation, etc.
Theory and evidence
Logical positivists made distinction between context of discovery and context of justification
Context of discovery: distinction should be drawn between the way in which a theory is discovered
Context of justification: rational reconstruction of theory according to tenets of logical positivism for
the purpose of its justification
Logical positivists argued that philosophy of science should only concern itself with the context of
justification. Context of discovery irrelevant to establishing scientific value of a theory (means: no
relevance to meaningfulness)
Another fundamental distinction by logical positivists: between theories & evidence, facts and data,
since theories depend upon latter for their justification
Scientific theories: systematic collections of concepts, principles, and explanations that organize our
empirical knowledge of the world. In advance of scientific knowledge: theory & evidence given
different weights and play different roles => main problem for philosophy of science & also economic
methodology based on logical positivists thinking, is to explain the relation between them.
Syntactic view: according to logical positivists understanding of theory, proper characterization of a
scientific theory consists of axiomatization in first-order formal language
First-order formal language consists only of symbols representing variables (x,y,x…), function symbols
(A(.), B(.), C(.),..), predicate symbols (A,B,C,…) and the – symbols (“not), v (“or”), A (“and”), ->
(“if…then”), A(upside down) “for all individuals, E (opposite) “for some individuals”
Axiomatization reduces a theory to a set of axioms in the manner of Euclid’s treatment of geometry.
Axioms are formulations of logical laws that specify relationships between theoretical terms. These
terms are implicitly defined by syntactics (=logical structure of theory). As result, language of theory
divided in two parts: (i) observation statements: (or “protocol sentence” that describe observable
objects or processes) (ii) theoretical statements
Meaning theoretical terms given by their observational consequences. Theoretical terms for which no
corresponding observational terms are regarded as meaningless
Theoretical terms identified w/their observational counterparts by means of correspondence rules
(=rules that specify admissible experimental procedures for applying theories to phenomena)
Operationalized: when theoretical terms defined completely in observational terms. According to
received view, for proposition to be meaningful necessary that all theoretical terms are
operationalized.
Example operationalization: “unemploymeny,” which is number of people who are jobless, looking for
jobs, and available for work. -> in this case easy to say who is jobless. However, definition of terms
“looking for jobs” and “available for work” involves human motivation -> cannot be observed as
physical fact, such as oranges being orange. Means, in order to measure unemploymeny, “looking for
jobs” and “available for work” need to be operationalized. This is done by defining “unemployed
persons” as:
<- according to this process,only at this stage of operationalization
can unemployment me measured.
Related to this idea of operationalization and emphasizing idea of meaningfulness, bridgeman
developed extreme form of empiricism (called operationalism). True meaning of a term is to be found
by observing what a scientist does with it, not by what a theory says it does or what its author thinks
, it does. Ways a term is measured. Consequence of this view: term will have different meaning when
it’s measured in a different way:
The nature of scientific explanation
explanation answer to a why question
received view of scientific explanation is more specific: should show some event or regularity to be an
instance of a fundamental law
Hempel developed this view systematically into deductive-nomological (DN) model of explanation (or
covering-law model)
In DN explanation, statement of what is to be explained (explanandum) is deducted from a set of true
statements that includes at least one law (nomos). This latter set of statements called the explanans:
- solid line represents deductive inference
- model explanation of, for ex: why firm x raised its price if one interprets symbols as follows:
- problem w/Hempel’s definition: according to this model, not all deductive inferences can be considered to be
explanations. Consider following ex:
- if assumed George is man, nobody would regard argument above as explaining why George does not get
pregnant. Doesn’t matter whether of not George took birth control pills
- in this instance taking pills not causally relevant for explanation of George not getting pregnant
- explaining law should identify causally relevant factors, but Hempel’s DN model, which focuses on (logical)
deduction from a law, does not require this.
There is an even more fundamental problem w/this model of explanation: required that
generalizations in the explanans be laws. How can we be certain this is the case?
-> closer consideration what laws are: generally suppose law is true but, for current discussion this is
not relevant -> refer to “law-like” statements as statements that are just like laws (don’t know if true).
Idea of “law-like” is just idea of regularity. Problem=many regularities appear to be accidental.
As a 1st attempt to distinguish law-like statements from accidental regularities, see whether a
syntactic definition of law-like statement might help. All law-like statements, then, have logical form
of universal condition:
Difference between two examples: accidental generalization contains designations for a particular
individual object and for a definite date or temporal period, while universal law does not. So
necessary requirement for universal law is that the domain of the antecedent is not restricted to
things falling into fixed spatial region or particular period of time. Law-like statements are required to
be unrestricted generalizations.
Another difference between two examples: implication arrow (->) in syntactic expression of universal
law above denotes a necessity. All elements in antecedent must necessarily also appear in
consequent. This is not the case for accidental generalization: a coin put in smith’s pocket during time
period a would not necessarily turn into a dime -> so law0like statement has no exception.
