Claude Monet: Gare St. Lazare, 1877
Claude Monet: Gare St. Lazare (1877).



In Sociology 202 we will learn the basic techniques of the qualitative and quantitative methods most commonly used by sociologists. For each method we will learn how it is applied, its strengths and weaknesses, and the scientific problems to which it is most often applied. Then we will learn the fundamentals of sociological data analysis and technical report writing. Before turning our attention to these topics we will learn the scientific context in which qualitative and quantitative methods are used. That is, we will learn the key principles of science.
Epistemologies
To understand sociology or any other science we need to understand the key principles of scientific inquiry. Before I describe these principles I will define science and compare it with four other epistemologies (ways of knowing about reality) using the typology (classification scheme) presented by Walter Wallace in The Logic of Science in Sociology. A scientific theory is a set of empirically falsifiable, abstract statements about reality. Simply put, it is a story about how reality works that can be falsified by observation.

Science requires theory for three reasons:
  1. Theory provides an explanation of why an event occurs. In contrast, empirical generalizations merely summarize a specific set of observations. Fishbein and Ajzen's theory of reasoned action, for example, is a set of abstract statements that can and have been applied successfully to understand and predict a very wide range of behaviors.
  2. Scientists use theory to help others in the community of scholars (persons trained and certified as members of a scientific discipline) with their investigations. Limitations to the theory of rational expectations discovered in a study of one behavior, for example, might prove helpful in understanding or predicting another behavior.
  3. By gaining support for theory (based upon analysis of quantitative data, qualitative data, or some combination of these), scientists feel confident about applying theory to improve the well-being of human, animal, and plant populations by building bridges, growing food, raising healthy families, and so on.
The ability of theories to be falsified by observation is the critical component of science that sets it apart from other forms of knowing.

To say that science necessarily entails the use of theory is not to say that science must be deductive (research designs that begin with an established theory). Quite to the contrary, an essential element of scientific inquiry is inductive creation/reformulation of theory. Still, it is the focus on theory, whether its testing through deductive procedures or its development through inductive procedures, that defines science as a unique epistemology.



Evolution or Creationism?

    To emphasize the importance of theory in science we can compare scientific and religious explanations of the origin of the species. Judeo-Christian stories about creation (i.e., Creationism, Intelligent Design), for example, which state that the universe and everything in it were created by a supernatural--and therefore unobservable--being, might in fact accurately depict creation, including the origin of the human species. These stories, however, cannot be falsified because one cannot disprove the existence of an absolute god or intelligent designer. Creationism and intelligent design, therefore, are not and can never be scientific theories. The theory of evolution, on the other hand, can be falsified by observation and thereby qualifies as a scientific theory.

    Of course, if one can never know Absolute Truth, one can never fully know that a theory has been falsified! That is, true falsification is never achievable (see this related article posted on Biocrawler)! Explanations of "what is science," therefore become complicated and compromised. As do attempts to distinguish science from "non-science," including attempts to dismiss creationism and intelligent design from the realm of science.

    Still, it is possible to draw a line in the sand between science and creationism/intelligent design because, in principle, one could not devise an experiment to test the existence of God or an Intelligent Designer but one could bring observations to bear on falsifying the theory of evolution. If you think the qualifier "in principle" is too big a concession to make for the purpose of defining science, then please recognize that if one attempts to make a logical argument for the existence of God then one must make big concessions also (see this related paper written by Wade A. Tisthammer).

    In summary, epistemologically, one cannot argue that creationism or evolution is a "better" or the "more correct" story. They simply are different types of stories. Science, however, MUST be based upon stories that in principle can be falsified by observation.

    The advantage of science over other approaches to knowing is that observations can be replicated by others using the same procedures that produced the original observations. Replication gives one a sense of confidence that an observation (e.g., the tensile strength of steel under certain conditions of temperature and pressure) did not occur by chance or miracle (e.g., leading one to have a certain amount of confidence that the bridge will not collapse).

    This PowerPoint presentation regarding ontology provides a more detailed comparision of science and intelligent design.


What is Good Science? To answer these questions, we must understand the rules of science. To understand the rules of science, we will trace the path of how the rules were developed by examining the philosophy of positivism and its various critiques by the philosophy of phenomenology.

Rules of Positivism

Positivism attempts to establish a set of rules for science that can verify the truthfulness of statements about 'reality' in an objective, value-free, unbiased manner. The positivist philosophy can be presented in various ways; the presentation below reduces positivism to four rules:
  1. Rule of operationalism: Record only that which is actually manifested in experience. Rely only upon sense data. Rule out the metaphysical or theological bases for verification. That is, only data directly observable by the senses are proper for scientific inquiry.

  2. Rule of nominalism: No generalized constructs or terms that cannot be reconstructed by reference to sense data.

  3. Rule of value-free knowledge: Scientific inquiry must be value-free and unbiased.

  4. Unity of scientific method The scientific method is universal and equally applicable to all areas of inquiry. All sciences must obey Rules 1-3 above.
The Phenomenological Critique of Positivism

Phenomenology argues that the rules of positivism, although noble in intent, are impossible to follow in practice. Blind adherence to the rules of positivism, argue phenomenologists, ignores the true nature and obscures the real value of science.
  1. Critique of the rule of operationalism: What constitutes sense data? How does one obtain pure sense data that is not filtered through the personality, experience, and preconceived ideas of the scientist? Given that humans are thinking beings, the rule of operationalism becomes not only restrictive to the social sciences, which seek to understand the thinking of individuals and collectivities, but in itself a contradiction in that scientists are thinking beings who make observations about reality. Nothing is observed directly by the senses; all observations are filtered through the experiences and biases of humans who interpret the raw sense data gathered by their eyes, ears, etc.

  2. Critique of the rule of nominalism: Logical atomism, or the reduction of all observations to their basic components of sense data results in an attempt to reduce all statements ad infinitum to some fundamental building block of reality. An understanding of reality, however, always reflects abstractions drawn from sense data. The sparrow, for example, might be reduced in description to the nature of its atomic structure. But the "sparrow" is an abstraction of these building blocks. All the description possible, from now until eternity, of the basic building blocks of the sparrow never will equal "sparrow" until the observer calls this collection of building blocks a sparrow, thereby creating the abstract concept: sparrow.

  3. Critique of the omission of values: Once the rules of operationalism and nominalism are shown to be logically impossible to follow, then it becomes evident that all observations are influenced by the values and biases of the observer. That is, observation is a human endeavor, one affected by values and bias.

  4. Critique of the principle of one science: If one cannot establish a set of rules for verification, then science becomes a human enterprise, subject to the dynamics of other human enterprises. Because no science can adhere to the rules of positivism, then none are required to do so. But this rule--that all sciences must adhere to the same set of guidelines--does hold true. All sciences--life, physical, and social--must adhere to the same rules. It is just that these rules cannot be the rules of positivism because the rules of positivism cannot be followed by any science.

The Hypothetico-Deductive Model

Another approach to verifying the truthfulness of statements about reality is to assess them as logical conclusions of laws established a priori through the human experience. The Hypothetico-Deductive (HD) model, in effect, admits that the rules of positivism are impossible to follow--that objective, value-free, unbiased observations are impossible to obtain. The HD approach is to establish a set of rules whereby objective, value-free, unbiased conclusions can be drawn from admittedly biased observations.

In the HD model, the explanandum (event to be explained) is a conclusion drawn from premises (explanans) that cover one or more universal laws.

The HD model takes the following form: For example: The Phenomenological Critique of the HD Model

The HD model allows for symmetry of explanation and prediction, but suffers from two fundamental problems:
The Community of Scholars Approach

The phenomenological critique of positivism refutes the principle of verification. So, the phenomenological approach to science is to relax the verification principle, but still rule out metaphysical justification (i.e., phenomenology retains the requirement of empirical falsification of statements about reality). This approach entails a big concession (i.e., that truth cannot be verified) and therefore requires establishing a criterion for deciding what constitutes sound science.

The phenomenological approach to evaluating science is to rely upon the consensus (or intersubjective) opinion of the community of scholars (i.e., basically, all those persons who hold a PhD degree in a particular scientific discipline) regarding the acceptability of statements about reality.

What is good science?

According to the phenomenological position, the answer to this question rests with the community of scholars. It is this community that decides when work meets the criteria of good science. And it is this community that decides not the truthfulness of statements, but their acceptance as the best statements possible until something better comes along.

In practice, technical reports of scientific investigation are submitted by the author(s) for review by the community of scholars (See for example, the Review of MS#03-27). Typically, the procedure is to submit a manuscript to a professional journal. The Editor of the journal distributes copies of the manuscript to 2-4 reviewers, who are not told the identity of the author(s). The reviewers evaluate the quality of the scientific investigation as it is reported in the manuscript. If they think the manuscript is clearly written and reflects acceptable scientific procedures, then they recommend it be published as an article in the journal. Upon publication, the study is considered to be "acceptable science."

The Phenomenological Critique of the Community of Scholars Approach

This approach of peer reviewing manuscripts for publication sounds straightforward, right?

Not so fast, argues Thomas Kuhn, in The Structure of Scientific Revolutions. Kuhn points out that the peer review process cannot be an objective one. It includes elements of other epistemologies, such as religious beliefs, authoritarianism, and mysticism. If the findings of an investigation challenge long-held beliefs, for example, they will be scrutinized more vigorously. If they challenge positions held by leading persons in the community of scholars or threaten strong economic benefits promised by a new technology, then they are looked upon with greater skepticism. If the findings do not sit well with the religious, political, or philosophical positions of the reviewers or the Editor of a professional journal, it will be more difficult for these persons to find the manuscript acceptable.

Thus, the community of scholars, like any other human collectivity, is influenced by power structures, economics, religion, politics, culture, and so on.

Summary

This page describes different types of epistemologies--ways of knowing. Science differs from other epistemologies in that its stories about reality must in principle be capable of being falsified by observation. Thus, science posits theories (stories about reality) that use abstract concepts (so they can be applied to many situations or events) to describe reality. These theories can be cast aside with sufficient evidence contradicting them.

Science is not necessarily a better epistemology than others, it is simply a different form of knowing. The advantage of science is that, with sufficient training, anyone can conduct scientific investigations. In principle, scientific findings are immune from special characteristics of the observer. Because scientific findings can be replicated by anyone with similar training and access to observations (e.g., equipment, funds, contacts with human or animal subjects), people gain a sense of confidence in scientific findings. Also, the peer review process helps to ensure that science is conducted with expertise and integrity.

But science is conducted by scientists, who exhibit individual traits and respond to the expectations of the collective. That is, science is influenced by politics, economics, religion, culture, and social relations. So, the enterprise of science includes other epistemologies, such as religion, authoritarianism, and mysticism. Scientists are committed to doing the best they can to behave in an objective, unbiased, and value-free manner. But they know that these goals cannot be reached. The philosopher P.W. Bridgman said it well, "The scientist has no other method than doing his damnedest."

Suggested Readings
These books and articles provide excellent summaries of the philosophy of science. They will reference philosophers who have made important contributions to understanding scientific inquiry.
  1. Benton, Ted (1977) Philosophical Foundations of the Three Sociologies. London: Routledge and Kegan Paul.
  2. Carmines, Edward G. and Richard A. Zeller (1979) Reliability and Validity Assessment. Beverly Hills, CA: Sage.
  3. Fales, Evan (1982) "Must Sociology be Qualitative." Qualitative Sociology 5(2): 89-105.
  4. Feyerabend, Paul (1975) Against Method. London, NLB.
  5. Giddens, Anthony (1974) Positivism and Sociology. London: Heinemann.
  6. Hamilton, Peter (1974) Knowledge and Social Structure. London: Routledge and Kegan Paul.
  7. Kuhn, Thomas S. (1962) The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
  8. Smart, Barry (1976) Sociology, Phenomenology and Marxian Analysis. London: Routledge and Kegan Paul.
  9. Wallace, Walter L. (1971) The Logic of Science in Sociology. New York: Aldine.
Links Related to the Philosophy of Science
Lyle Zynda's lectures on the philosophy of science delve perhaps too deeply into some topics than is necessary for this course, but provide a good background on the key issues affecting scientific inquiry.

Wade A. Tisthammer's paper on The Nature and Philosophy of Science addresses many of the same topics covered on this page.

Patrick O'Driscoll provides a comprehensive explanations of logical fallacies at Fallacy Files.
Link to Home Page Link to Syllabus Link to Class Assignments Link to Readings