Malhotra, Yogesh. (1994).

On Science, Scientific Method And Evolution Of Scientific Thought: A Philosophy Of Science Perspective Of Quasi-
Experimentation [WWW document]. URL http://www.brint.com/papers/science.htm

Evolution of Scientific Thought - Positivism to Scientific Realism

Logical Positivism
During much of this century "positivism" has dominated discussions of scientific method. The term was
popularized by Comte, and generally refers to a strict empiricism which recognizes as valid only
knowledge claims based on experience (Abbagnano, 1967; Brown 1977).
During the 1920s positivism emerged as a full-fledged philosophy of science in the form of logical
positivism. Developed by the Vienna Circle, a group of scientists and philosophers, logical positivism
accepted as its central doctrine Wittgenstein's verification theory of meaning (Brown, 1977; Passmore,
1967). The verification theory holds that statements or propositions are meaningful only if they can be
empirically verified. This criterion was adopted in an attempt to differentiate scientific (meaningful)
statements from purely metaphysical (meaningless) statements (Anderson, 1983).
According to logical positivists, universal scientific propositions are true according to whether they have
been verified by empirical tests -- yet no finite number of empirical tests can ever guarantee the truth of
universal statements (Black, 1967; Brown, 1977; Chalmers, 1976). In short, inductive inference can never
be justified on purely logical grounds (Hempel 1965). As a result of these difficulties, Carnap (1936,
1937) developed a more moderate version of positivism, which has come to be known as logical
empiricism which became the "received view" in the philosophy of science for approximately next 20
years (Suppe 1974).

Logical Empiricism
Essentially, Carnap replaced the concept of verification with the idea of "gradually increasing
confirmation" (1953, p. 48). He argued that if verification is taken to mean the "complete and definitive
establishment of truth," then universal statements can never be verified. However, they may be
"confirmed" by the accumulation of successful empirical tests. Thus, science progresses through the
accumulation of multiple confirming instances obtained under a wide variety of circumstances and
conditions. Logical empiricists believe that all knowledge begins with observation. This leads to
empirical generalizations among observable entities. As our ideas progress, theories are formulated
deductively to explain the generalizations, and new evidence is required to confirm or disconfirm the
theories. Throughout the process, data are given precedence. Indeed, the entire process is viewed as
essentially an inductive one. Science in general and knowledge in particular are believed to occur in an
upward fashion: from data to theory to understanding (Bagozzi, 1984). Feigl (1970: p. 7) terms this as "an
'upward seepage' of meaning from the observational terms to the theoretical concepts," and it is construed
in a similar way by Hempel (1952: p. 36), Carnap (1939: p. 65) and others logical empiricists.
Logical empiricism is characterized by the inductive statistical method. In this view, science begins with
observation, and its theories are ultimately justified by the accumulation of further observations, which
provide probabilistic support for its conclusion. Of course, the logical empiricist's use of a probabilistic
linkage between the explanans and the explanandum does not avoid the problem of induction. It remains
to be shown how a finite number of observations can lead to the logical conclusion that a universal
statement is "probably true" (Black, 1967). Moreover, attempts to justify induction on the basis of
experience are necessary circular. The argument that induction has worked successfully in the past is itself
an inductive argument and cannot be used to support the principle of induction (Chalmers, 1976).
In addition to the problem of induction, logical empiricism encounters further difficulties because of its
insistence that science rests on a secure observational base. There are at least two problems here
(Anderson, 1983). The first is that observations are always subject to measurement error. The second, and
perhaps more significant, problem concerns the theory dependence of observation. We have discussed
some aspects of this issue under the section on Knowledge and Objectivity. The fact that observation is
theory laden does not, by itself, refute the logical empiricist position. It does, however, call into question
the claim that science is securely anchored by the objective observation of "reality."
In his development of falsificationism, Popper has offered an alternative method of theory justification
which is designed to overcome some of the difficulties inherent in logical empiricism.

Popper And Falsificationism

West Island School – TOK – SJT The Scientific Method – 1 - Development

Malhotra, Yogesh. (1994). On Science, Scientific Method And Evolution Of Scientific Thought: A Philosophy Of Science Perspective Of Quasi-
Experimentation [WWW document]. URL http://www.brint.com/papers/science.htm

Unlike positivists, Popper accepted the fact that "observation always presupposes the existence of some
system of expectations" (1972: p. 344). For Popper, the scientific process begins when observations clash
with existing theories or preconceptions. To solve this scientific problem, a theory is proposed and the
logical consequences of the theory (hypotheses) are subjected to rigorous empirical tests. The objective of
testing is the refutation of the hypothesis. When a theory's predictions are falsified, it is to be ruthlessly
rejected. Those theories that survive falsification are said to be corroborated and tentatively accepted
(Anderson, 1983).
In contrast to the gradually increasing confirmation of induction, falsificationism substitutes the
logical necessity of deduction. Popper exploits the fact that a universal hypothesis can be falsified by a
single negative instance (Chalmers, 1976). In Popper's approach, if the deductively derived hypotheses are
shown to be false, the theory itself is taken to be false. Thus the problem of induction is seemingly
avoided by denying that science rests on inductive inference. Anderson (1983) notes that Popper's notion
of corroboration itself depends on an inductive inference. According to falsificationism, then, science
progresses by a process of "conjectures and refutations" (Popper 1962, p. 46). In this perspective, the
objective of science is to solve problems.
Despite the apparent conformity of much scientific practice with the falsificationist account,
serious problems remain with Popper's version of the scientific method. For example, Duhem (1953) has
noted that it is impossible to conclusively refute a theory because realistic test situations depend on much
more than just the theory that is under investigation. Quine-Duhem thesis (Quine, 1953; Duhem, 1962)
points out that because of all of the background assumptions that might be wrong -- flaws in the
equipment, the effects of unknown or wrongly disregarded physical processes, and the like -- any outcome
can be rationally distrusted and explained away by ad hoc hypotheses that alter the background
assumptions. Falsification can thus be regarded as particularly equivocal (Cook and Campbell, 1979). The
recognition that established theories often resist refutation by anomalies while new theories frequently
progress despite their empirical failures, led a number of writers in the 1950s to challenge the positivistic
views of Popper and the logical empiricists (Suppe 1974). Various philosophers and historians noted that
scientific practice is often governed by a conceptual framework or world view that is highly resistant to
change. In particular, Kuhn pointed out that the established framework is rarely, if ever, overturned by a
single anomaly (1962). Kuhn's model helped to initiate a new approach in the philosophy of science in
which emphasis is placed on the conceptual frameworks that guide research activities.

Kuhn’s Scientific Revolutions

Central to the Kuhnian argument is the concept of a "paradigm." A paradigm constitutes the world view of
a scientific community (Laudan, 1977; Suppe, 1974). The paradigm will include a number of specific
theories which depend, in part, on the shared metaphysical beliefs of the community (Kuhn, 1970). In
Kuhn's view, the individual scientist's decision to pursue a new paradigm must be made on faith in its
"future promise" (Kuhn 1970: p. 158). Furthermore, in his view, science progresses through "paradigm
shifts," but there is no guarantee that it progresses toward anything -- least of all toward "the truth" (Kuhn
1970, p. 170). Given its (seeming) advocacy of relativism, Kuhn's Structure of Scientific Revolutions
became one of the most carefully analyzed and evaluated works in the philosophy of science. [Relativism
is discussed further in the paper.] In criticism of Kuhn, some writers (such as Lakatos, 1974; Laudan,
1977) have suggested alternative world view models. Here we briefly explain Laudan's "research
tradition" concept, which attempts to restore rationality to theory selection by expanding the concept of
rationality itself.

Research Traditions
Like Kuhn and Lakatos, Laudan sees science operating within a conceptual framework that he calls a
research tradition (Anderson, 1983). The research tradition consists of a number of specific theories, along
with a set of metaphysical and conceptual assumptions that are shared by those scientists who adhere to
the tradition. A major function of the research tradition is to provide a set of methodological and
philosophical guidelines for the further development of the tradition (Anderson, 1982). Following both
Kuhn and Popper, Laudan argues that the objective of science is to solve problems -- that is to provide
"acceptable answers to interesting questions" (Laudan, 1977, p. 13). On this view, the "truth" or "falsity"
of a theory is irrelevant as an appraisal criterion. The key question is whether the theory offers an
West Island School – TOK – SJT The Scientific Method – 1 - Development
Malhotra, Yogesh. (1994). On Science, Scientific Method And Evolution Of Scientific Thought: A Philosophy Of Science Perspective Of Quasi-
Experimentation [WWW document]. URL http://www.brint.com/papers/science.htm

explanation for problems that arise when we encounter something in the natural or social environment
which clashes with our preconceived notions or which is otherwise in need of explanation .

Critical Relativism
Critical relativism is a multifaceted philosophy of science: one of its major assertions is that there exists
no single "scientific method." Instead, disciplinary knowledge claims are viewed as contingent upon the
particular beliefs, values, standards, methods, and cognitive aims of its practitioners. Moreover, critical
relativism recognizes that knowledge production in the social sciences is impacted by the broader cultural
milieu in which it is embedded (Anderson, 1986).
Critical relativism is skeptical of all claims to scientific knowledge because it recognizes that there
are multiple scientific objectives and alternative methods for attaining these objectives (Laudan, 1984).
Moreover, critical relativism recognizes that the value of such claims must be assessed in light of their
unique modes of production and their methods of justification. To suggest that the hallmark of scientific
knowledge is its empirical testability is to settle for far less than we should demand of such an important
enterprise as science. Anderson (1986) further argued that the requirement of empirical testability is
notoriously ambiguous within the recognized sciences, and it is a criterion that is allegedly met by
patently "nonscientific" disciplines (Laudan, 1983).
Critical relativism rejects the basic premise of the positivistic approaches that there is a single
knowable reality waiting "out there" to be discovered via the scientific method (Olson 1981). The critical
relativist has no quarrel with the metaphysical notion that there may well be a single social and natural
reality, but he or she will resist the assertion that science is capable of revealing or even converging upon
this "reality" (Laudan, 1981). Instead, the relativist accepts competing research programs for what they are
-- different ways of exploring and analyzing natural phenomena, each with its own advantages and
liabilities (Anderson, 1986).

Scientific Realism
After its brief excursion into the relativism, constructivism, and irrationalism of Kuhn and Feyerabend in
the 1960s, philosophy of science turned toward realism in the 1970s (Suppe 1977). In other words, the
reasoned pursuit of truth returned to the philosophy of science.
Classical realism believes that the world exists independently of its being perceived (Hunt, 1990). A
fundamental tenet of modern-day, scientific realism is the classical realist view that the world exists
independently of its being perceived. This is contra Olson's (1981) relativism: there really is something
"out there" for science to theorize about (Hunt, 1990). However, scientific realism does not embrace
"direct" realism which holds that our perceptual processes result in a direct awareness of or
straightforward confrontation with objects in the external world. Advocates of scientific realism, though
agreeing that our perceptual processes can yield genuine knowledge about an external world, emphatically
reject direct realism. They argue for a fallibilistic and critical realism. Hence scientific realism is a middle-
ground position between direct realism and relativism. Scientific realism is also a critical realism,
contending that the job of science is to use its method to improve our perceptual (measurement) processes,
separate illusion from reality, and thereby generate the most accurate possible description and
understanding of the world (Hunt, 1990). The practice of developing multiple measures of constructs and
testing them in multiple contexts in social science stems from this critical orientation (Cook and
Campbell, 1986). In short, scientific realism proposes that (1) the world exists independently of its being
perceived (classical realism), (2) the job of science is to develop genuine knowledge about the world, even
though such knowledge will never be known with certainty (fallibilistic realism), and (3) all knowledge
claims must be critically evaluated and tested to determine the extent to which they do, or do not, truly
represent or correspond to that world (critical realism). In conclusion, with respect to truth and scientific
realism, the perspective of Siegel (1983, p. 82) seems a fair summary statement: "To claim that a scientific
proposition is true is not to claim that it is certain; rather, it is to claim that the world is as the proposition
says it is."

West Island School – TOK – SJT The Scientific Method – 1 - Development