scientists and philosophers throughout the 20th century. Its premises and

practices were most clearly and persuasively stated by two

philosophers—John Dewey and Karl Popper. Independently, they defined the school

quite similarly, but their judgments of it were irreconcilable.

Dewey was a practitioner of instrumentalism who, while fearing that

the name was easily misunderstood, adopted it for his modernization of

tools of induction and his denial of reality behind experience. Popper was a

critic who judged its insistence on induction and its denial of reality

behind experience to be hopelessly flawed. These contrary judgments endowed

the school with the legacy of confusion and ambiguity described below.

This article reports the definition of instrumentalism accepted by these two

philosophers. It explains the grounds of their irreconcilable judgments, now

embedded in popular understanding of the school. And it describes the practice of

followers of each philosopher, demonstrating that neither philosopher's

judgments have achieved universal assent, leaving the school's meaning and

legitimacy in modern scientific inquiry indeterminate.

Instrumentalism defined In 1925, John Dewey published an article

entitled "The Development of American Pragmatism," in which he defined

instrumentalism to distinguish it from schools known as "pragmatism" and

"experimentalism." In 1956, Karl Popper published an article entitled "Three

Views Concerning Human Knowledge," in which he defined instrumentalism to

distinguish it from "essentialism" and a "third view"—his own-which he came to

call "critical rationalism." Dewey's article was republished in 1984

in John Dewey: The Later Works. Popper's article was republished in 1962 in

Conjectures and Refutations. The following four premises defining

instrumentalism are taken from these sources. Premises 1 and 2 were accepted

by both philosophers and the general public. Premises 3 and 4 were and remain

controversial, and will be analyzed in the next section.

1) Theories are instruments, tools-of-the-trade of thinking.

Dewey: Popper:

2) Theories are tested by consequences, applying the instrumental criterion of

judgment. Dewey:

Popper: 3) Theory-development requires inductive

reasoning, basing general statements on limited observations.

Dewey: Popper:

4) There are no realities beyond what can be known using instrumental

theories. Dewey:

Popper: Instrumentalism is often identified with

other schools which share some of these premises: positivism, pragmatism,

operationalism, behaviorism, anti-realism, empiricism,

consequentialism. Instrumentalism judged

Dewey and Popper disagreed on premises 3 and 4. The primary grounds of their

disagreement were expounded in the 1930s. In 1935, Popper published Logic

of Scientific Discovery, in which he used traditional logical forms to

criticize modern schools of thought, including instrumentalism. In 1938,

Dewey published Logic: the Theory of Inquiry, in which he reconstructed

traditional logical forms to make them usable by modern schools of thought.

Neither of these volumes used the name instrumentalism, but both discussed and

judged the premises above. = Popper's critique=

The opening paragraph of Popper's The Logic of Scientific Discovery observed

that all modern empirical schools accept premises 1 and 2, which he later

identified with Instrumentalism: Several paragraphs later, he admitted

the popularity of induction—premise 3—but denied its capacity to generate

logically true theories: Popper's reference to swans recalls a

famous historic error: the inductively-derived belief that all

swans are white. He labelled the practice illogical: "Now in my view

there is no such thing as induction. Thus, inference to theories, from

singular statements which are 'verified by experience', is logically

inadmissible." Popper rejected inductive reasoning in

favor of deductive reasoning because he maintained that the former could not

achieve logical form. Deduction can move from a self-evident universal statement,

such as "All men are mortal", to true singular statements that every

individual human is mortal, because the universal statement already embraces all

singulars. But there can be no principle by which a singular statement can

justify a universal, because no singular statement can report observing "all" of

any kind. Popper rejected induction—premise 3–-but

not premise 2—the criterion of instrumental efficiency. He argued that

deduction could serve modern science, not by assuming general statements to be

true, but by providing general statements testable by their

consequences. Falsification "works" when experience contradicts a theory's

predictions: "it must be possible for an empirical scientific system to be

refuted by experience." Popper rejected premise 4 because it

denies the distinction between pure and applied science. He granted that science

might be viewed from an empirical or instrumental point of view, but asserted

that an epistemological or reality point-of-view was equally valid, meaning

logical truths can be found independently of experience. His

evidence was that pure sciences such as mathematics and logic can make true

statements without observing facts-of-the-case.

Logically true theories don't require establishing facts-of-the-case; they can

be conjectural myths, derived from inspiration or chance, which are "...

psychologically or genetically a priori, i.e., prior to all observational

experience." They can also precede observation or recognition of

similarities and differences. Instrumentalism's denial of

logically-certain deductive truths—premise 4—threatens "... the idea

of the objectivity of knowledge and of common standards of criticism or

rationality." Because Instrumentalists claim that "truth" is always

situational, they forfeit their capacity to explain sciences in which the

instrumental criterion of judgment cannot be applied. In pure sciences, the

criterion is logically-established truth, not what works or is useful given

temporary conditions. = Dewey's reconstruction=

Dewey's Logic of 1938 was very different from Popper's Logic of 1935. While

Popper used traditional logical forms to criticize modern practices such as

induction, Dewey revised those forms. He addressed the problem of whether

scientific inquiry "can develop in its own ongoing course the logical standards

and forms to which further inquiry shall submit." His affirmative answer is the

substance of premise 4, which traditional logic led Popper to deny.

Dewey's Logic did not name instrumentalism or pragmatism, but

asserted that both schools treat theories as tools for producing

consequences. Consequences are "necessary tests of the validity of

propositions, provided these consequences are operationally

instituted and are such as to resolve the specific problem evoking the

operations, ..." When Dewey analyzed induction—premise

3—he accepted its standard meaning of processes for developing general

propositions from particular cases. He explained why Aristotle's understanding

of these methods was no longer acceptable.

Popper partially repudiated Aristotle's belief that superior intellects can

"intuit the essence" of eternal forms by observing physically changing forms:

each observed swan is an imperfect sample of universal-but-unobservable

swan-ness. But he provided no rational means to carry out induction's necessary

function of establishing the facts-of-the-case by relating singular

observations of kinds to general statements about kinds. Dewey's

instrumental analysis did provide such means by reconstructing both induction

and deduction. One may think of a singular observation,

i.e., "this swan is white", as an isolated fact without general reference.

But Dewey insisted that such an observation necessarily involves the

general meaning of "swan" as a particular kind of "bird". If one were

not familiar with a kind of animal having numerous well-established

characteristics, one could not name it either "bird" or "swan". Kinds,

including species, do not exist apart from experience. They are created by

inquiries which—contrary to Popper—use induction to distinguish stable

characteristics of experience from accidental or irrelevant

characteristics. Dewey argued that modern science does

not treat particular observations as knowledge of what is real: one does not

assume, after a few observations, that whiteness is a defining characteristic

of swans. Particular observations "are selectively discriminated so as to

determine a problem whose nature is such as to indicate possible modes of

solution." Observations become facts-of-the-case only after being

causally related to a problem. Dewey supported this theoretical

generalization with an example of medical knowledge. The case of malaria

shows how modern induction avoids Popper's charge of requiring endless

observations. After certain "singular" symptoms came

to be recognized as constituting a disease, it was named malaria—literally

"bad air"—as a common-sense conjecture about its cause—premise 1. Popper might

have considered that conjecture to be testable by predicting that the disease

would be absent in environments with "good air." But testing a prediction

about air quality could not have led to new insights. It was an insignificant

fact-of-the-case. When further observations—applying

premise 3—identified the conjunction of parasites with the disease, experiments

revealed the life-history of particular parasites and their relation to a

particular kind of mosquito: anopheles. At each stage of inquiry, particular

observations [inductions] led to general hypotheses [deductions] guiding further

observations to establish logically-warranted particular and

general propositions. Multiple theories generated by induction were used

throughout the process of inquiry. They evolved from quite conjectural to quite

confirmed generalizations, but never from "conjectural myths" to "truths"

independent of observable life processes.

The result of this hypothetical-deductive sequence was to

establish malaria as a specific kind of disease with a determinate etiology.

Dewey affirmed the logical force of this demonstration. It provides the logical

principle justifying induction, the possibility of which Popper denied.

With this logical principle, Dewey validated induction—premise 3—as well as

his rejection of realms such as pure science capable of establishing

objective truths unknowable by applied science—premise 4. He argued that

warranted generalizations never exist apart from experience. They arise only

in the process of inquiry, making invalid any claim to truths "logically

prior to observation or recognition of similarities and differences."

But the dependence of warranted theories on situational factors—induction—does

not eliminate objective standards of judgment, as Popper feared. Both ends

and means have consequences that can be judged more or less instrumentally

efficient—premise 2. In summary, Dewey's reconstruction of

logic directly refuted Popper's argument for rejecting induction and for

maintaining the distinction between pure and applied science. His instrumentalism

requires hypothetical-deductive operations to establish warranted

assertions to solve problems—employing all four premises.

Instrumentalism practiced Dewey and Popper never confronted their

differences. Consequently, the advocate's and the critic's

irreconcilable patterns of thought remain identified with the school.

Current use of the name embraces this incoherent legacy.

To exemplify this continuing ambiguity, this article examines recent practice by

significant people influenced by each philosopher's view of instrumentalism.

Economist Milton Friedman identified himself with the theory and practice of

Popper, while philosopher Larry Hickman and economist John Fagg Foster

identified themselves with Dewey. Should any of them be called adherents of

instrumentalism? = Milton Friedman's practice of

Instrumentalism= Milton Friedman was a Nobel laureate in

economics who contributed to the two branches into which economics is often

divided: a pure value-free science—positive economics—and an

applied normative science—political economy. He participated in the Mount

Pelerin Society to which Karl Popper belonged. In 1953 he published an

essay—"The Methodology of Positive Economics"—which came to identify him

with instrumentalism despite never mentioning that school, or Popper, or

Dewey. Friedman explicitly embraced premises 1

and 2 when he identified the task of positive economics as providing "a

system of generalizations or conjectures that can be used to make correct

predictions about the consequences of any change in circumstances." But his

position on premises 3 and 4 was ambiguous. Contrary to Popper, he

appeared to approve of basing theoretical conjectures on

facts-of-the-case provided by inductive observations—premise 3:

But he joined Popper in rejecting premise 4—that conjectures must derive

from descriptively true assumptions. This rejection appears to make

irrelevant the practice of relating theories to facts by induction.

In words close to Popper's praise of false conjectures, Friedman praised

purely mental hypotheses derived from inaccurate assumptions:

Friedman's 1953 essay provoked extensive criticism from both orthodox and

heterodox economists. In 1959, economist Lawrence Boland published "A Critique of

Friedman's Critics", in which he asserted that all critics were wrong

because they failed to understand that Friedman was an Instrumentalist.

The "coherent philosophy" which Boland identified with approval as

instrumentalism included premises 1 and 2, acceptable to both Popper and

Dewey—using theories as instruments to generate successful predictions. But

Boland left out of his definition premises 3 and 4—the premises Popper

rejected along with the name. Because Friedman downplayed inductive operations

and praised unrealistic hypotheses, Boland felt justified in praising him as

an instrumentalist, although the same logic would justify praising Popper as

an instrumentalist. Boland's paper generated further debate

over the meaning of instrumentalism and whether the school Popper rejected could

be made acceptable. In 1989, economists Abraham Hirsch and Neil De Marchi

published a detailed analysis of Friedman's professional work, which

found Friedman's practice inconsistent and Boland's interpretation misleading.

After analyzing Friedman's theoretical and practical writings, Hirsch and De

Marchi reached convoluted conclusions. They agreed that Friedman sometimes

practiced what Boland called instrumentalism, applying premises 1 and

2. But they also found much of his work compatible with Dewey's instrumentalism

but not Popper's—applying premises 3 and 4.

Hirsch and De Marchi recognized the irreconcilability of Popper's "notions

of deductive explanations" which avoid induction and Dewey's "process-view of

inquiry" which requires both induction and deduction. They concluded that these