links students ito uchii newsletters gallery English index index

The following is the webpage for Merrilee Salmon's course on the philosophy of social sciences, given at Kyoto University, Spring 2000.

The page was prepared by Soshichi Uchii, who, as the host, assisted Salmon's course.

back to Uchii Index


Philosophy of Social Science

M. Salmon

Outline of first lecture

Problem: Can there be a social science on the model of the natural sciences?

1. What are the "social sciences"?

2. What makes a discipline a science?

3. What makes the study of humans prima facie different from the study of other parts of the natural world?

(Special properties of humans that might interfere with studying humans scientifically)

4. Some scientific features of the study of human behavior.

5. Is the study of humans scientifically inferior to the study of the physical world?

6. If the answer to the previous question is yes, what can be done about it?


Outline of second lecture

(Continuation of the discussion of whether there can be a social science on the model of the natural sciences).

1. Naturalism presents an affirmative answer to the question above.

2. Exponents of naturalism: John Stuart Mill (late 19th century); C.G. Hempel (20th century. (Hempel's work will be discussed in Lectures 3 and 4 (second week of classes).

3. Mill's Methods for discovering causes and justifying causal claims: Method of agreement; Method of difference; Joint method; Method of concomitant variation; Method of residues.

4. Using the methods to find regularities in the social world. Historical investigations, study of character, laws of the mind.

5. Noncausal regularities vs. causal regularities, mechanisms, etc.

6. Criticism of Naturalism: ignores meaning of human behavior; skepticism about finding laws; critical theory, sociology of knowledge.

7. What is needed for an adequate philosophy of social science? (Fay and Moon)


Outline of third lecture

Review: Mill's regularity view of laws, versus a realist conception.

1. Continued discussion of Naturalism; Hempel's claim that explanation in science and in history are structurally similar to one another. According to Hempel, any adequate or good explanation of an event provides grounds for expecting that the event to be explained would occur. These grounds include descriptions of initial determining conditions (circumstances pertaining at particular times and places) and statements of general laws, which connect the initial conditions with events of the type to be explained.

Examples:

(1) Event to be explained: The recent volcanic eruption on Hokkaido Island.

Initial conditions: Buildup of subterranean gases and liquids under pressure at high temperatures; instability of surface of volcanic mountain; and so forth.

General laws: Laws of physics and geology connecting volcanic eruptions with the types of subterranean forces and instabilities described in the initial conditions.

(2) Event to be explained: Nian rebellion (a mid-nineteenth century peasant rebellion in North China). (E. Perry Rebels and Revolutionaries in North China 1845-1945.)

Initial conditions: Precarious ecology; existence of predatory strategies for survival (smuggling, petty theft, banditry); existence of protective strategies as well (local militias, fortifications).

General laws: As struggle to survive becomes more difficult, more peasants join bandit gangs. When bandit groups become large, they become targets of state repression. When groups are targets of state repression, they use whatever organizational means they have to fight back.

Summary: Explanations are arguments to the effect that the event to be explained was to be expected on the basis of the explanatory premises. The argument must be deductively valid or inductively strong. The premises must state initial conditions and include at least one general law. The initial conditions and (at least one) general law must be empirically verifiable.

2. Explanation sketches

3. Explanation as empathic understanding

4. Genetic (developmental) explanation

5. Explanation by motivating reasons

Wm. Dray denies laws are needed: To explain why Agent A does X, show that A was Also note, possible explanations in terms of character traits, or emotional states, etc. in a situation of type C, and that in a situation of type C, the appropriate thing to do is X.

Hempel's proposal

a. A was in a situation of type C.

b. A was disposed to act rationally.

c. Any person who is disposed to act rationally will, when in situation of type C, do X.

Also note, possible explanations in terms of character traits, or emotional states, etc.


Outline of fourth lecture

Problem: Is explanation in the social sciences structurally similar to explanation in the physical sciences? Are laws required?

1. Hempel proposes a schematic law for explanations by motivating reasons: "Any person who is disposed to act rationally will, when in a situation of type C, invariably (with high probability) do X."

Hempel defends this law against the charge that it is not empirical: "Dispositional concepts of the kind invoked in our explanations have to be regarded as governed by entire clusters of general statements‹we might call them symptom statements‹which connect the given disposition with various specific manifestations, or symptoms, of its presence (each symptom will be a particular mode of "responding" or acting under specified 'stimulus' conditions) and the whole cluster of these symptom statements for a given disposition will have implications which are plainly not analytic. ... Under these circumstances it would be arbitrary to attribute to some of the symptom statements the analytic character of partial definitions." Explanation in science and in history, p.28.

2. A contemporary development of Hempelıs proposal: the decision theoretical approach to explaining human behavior. (D. Little, Varieties of Social Explanation1991; J. Elster,The nature and scope of rational-choice explanation (text, Ch. 20); D. Papineau, For Science in the Social Sciences, 1988.

Example: Decision to buy a lottery ticket that costs 100 yen. Various outcomes, each associated with a probability and a value (utility). Matrix:

Actions

Outcomes

Expec. Val.

Win

Lose

@

Buy

Pr= 0.001 Val= 5,0000-100

Pr= 0.999; Val = -100

50-100

Don't buy

Pr = 0

Val= 0

Pr =1; Val =0

0

@

Remarks about the example, assignments of values not always monetary, etc.

To adapt this to Hempelıs model of explanation, consider an agent's assignment of probabilities to beliefs about outcomes in the world, and, the agent's assignment of value to those outcomes. These are the initial conditions. The "law" is that rational agents always choose an action that maximizes expected value. Value of this approach: a reconciliation of interpretivist and naturalist positions. Problem: Definition of rational agent? Lack of empirical content of law.

3. Not all explanations in the social sciences are explanations that refer to human decisions, choices. Examples: Murdock's "structural" explanation of forms of marriage in terms of form s of kinship; Explanations of how hierarchically structured societies (chiefdoms and more complex organizations) develop from simple egalitarian societies.

4. Micro explanations of macro social phenomena. Mechanism of rational choice


Outline of fifth lecture

Topic: Kincaid's defense of the claim that there are laws in the social sciences.

1. Criticism: Macro laws in social science require micro foundations in terms of individual behavior (such as those supplied by rational-choice theorists).

2. Are laws inadequate for explanation without microfoundations? Mill , Hempel, van Fraassen. Durkheim: social facts require social causes for their explanation. But, consider: "The crime rate rises in high-rise buildings with the height of a building up to a height of thirteen floors, but at more than thirteen floors levels off" O. Newman.

3. Can purely social laws be confirmed? How are laws (or empirical generalizations) confirmed? Eliminate spurious correlation. For causal claims, this means (1) eliminate coincidental connections, (2)eliminate correlations that occur as a result of an underlying common cause; (3)eliminate correlations that do not exhibit the proper temporal order.

Examples: Durkheim investigated factors such as climate, form of government, financial condition, marital status, and so forth as possible causes of higher rates of suicide. Neighborhood in which high-rises occur is a possible cause of higher crime rate.

4. Not all social laws are causal, however, and (2) and (3) above do not apply in the case of noncausal laws.

5. Even when laws are causal, we need not identify the causal mechanism to confirm them, but just be sure that we have controlled for possible spurious causes. Kincaid claims that biology is able to establish laws relating changes in gene frequencies to environmental changes without knowing the genetic mechanism involved (selection for or selection of).

6. Problem of reaching the end of chain of causal mechanisms.

7. Problem: Many social laws are teleological laws. How can we confirm that some feature exists in order to bring about some result (Example: In Andaman society the joking relationship between mother-in-law and son-in-law (A) exists in order to control social conflict (B).)

8. Kincaid (following Larry Wright) 1. Show that A has effect B; and 2. Show that having effect B plays a positive causal role in the existence, or continuing existence of A.

How to show 2.? Answer: Eliminate other possible causal factors, see whether A ever persists without having effect B, and so forth (follow the usual procedure for testing causal claims).

9. Problems with proposed "laws" of social science:

10. Confirmed laws in social sciences:

Economics: Rise in price-> decrease in demand

Decline in supply -> rise in price Assumptions?.

Ceteris paribus clauses? Transitivity of preferences. Resilience of laws even when transitivity is not present. Examples from ecology and evolutionary biology a good model for social science laws, for example, laws of development from simple to complex social arrangements.

11. How to explain lack of success in social science? No success? Slow progress?


Outline of Sixth Lecture

Topic: The Interpretivist Alternative to Naturalism: Interpretation, Understanding, Hermeneutics.

1.Historical remarks. Influences: Biblical exegesis; Literary interpretation; W. Dilthey (1833-1911) ; A. Schutz (1899-1959--phenomenology); R.G. Collingwood (1889-1943); Later Wittgenstein (P. Winch).

2. Features of interpretivist approach:

3. Collingwood

4. Subject matter of history: Events with an inside as well as an outside. The inside, or intention, of the agent makes the act the kind of act it is, and is logically inseparable from it.

5. Hard sayings. "All history is the history of human affairs." "All history is the history of thought." "The history of thought, and therefore all history, is the reenactment of past thought in the historianıs own mind."

6. Critical activity of the historian. Use of historical materials, written and unwritten. Relationship to the hypothetico-deductive method of confirmation.

7. Extensions of Collingwood's views: W. Dray; P. Winch


Outline of Lecture 7

Topic: Continuation of the discussion of interpretivism.

1. Collingwood asks why a science of human nature, which would give the human mind knowledge of itself, has failed. His answer: " its method has been distorted by the analogy of the natural sciences." What is that method? (Mill? Observing regularities, forming empirical generalizations, deriving those from higher-level generalizations, etc.)

2. Collingwood: "It is evident that such a science of human nature... could hope for results of extreme importance. " (Applications to the problems of moral and political life, comparisons to theoretical physics and its applications) p. 164

3. What constitutes success--and successful applications--of a science of human nature?

4. Collingwood's method for constructing a science of human nature: the plain historical method. "Reconstruction of [the entire human past] from documents written and unwritten, critically analyzed and interpreted." P.165 "History is what the science of human nature professed to be."

5. Comparison to hypothetico-deductive method.

What is the content of the historian's hypothesis? Must hypothesis be about or include meanings of events studied. How are these hypotheses tested? What is criterion for successful hypothesis?

6. Winch (The Idea of a Social Science) amplifies Collingwood's insight that knowledge of the meaning of an action requires not only knowledge of the agent's thoughts, but also the rules or form of life that constitute the agent's actions.

Relationship to Wittgenstein's views on language. Private language argument. Criterion for what sort of action an action is depends on social rules. Criterion for whether two actions are of the same kind also depend on the rules of that type of activity: Religious examples (praying): standing in the front of main hall of temple, pressing palms together; standing in front of a shrine, after washing hands and mouth with pure water and tossing a coin into a box, pulling a rope, bowing twice, clapping loudly twice, bowing again twice, and then stepping back and to the side.

Social examples: forms of politeness, ritualized expressions.

7. Implications for studying human behavior: Winch (following Collingwood) says this is more like a philosophical (epistemological) activity than an empirical study. Predictions possible. Basis for predictions: understanding of the rules. Meaning of failed prediction different in physical and social sciences.

8. Implications of importance of rules for clarifying "empathic understanding."


Outline of 8th lecture

Topic: Taylor on hermeneutic interpretation. Remarks on social context in which this work was produced.

1. 3 requirements for a field of study to be the object of hermeneutics:

2. Importance of these three criteria.

3. Subjectivity and the hermeneutic circle. The search for certainty.

4. Taylor's account of logical empiricism as an epistemological theory. Verification; Brute data must be separated from logical inferences (or computations, inductive or deductive) based upon those data. He claims---and will try to defend his claim---that this method is sterile as far as human science is concerned.

5. Taylor tries to show that human sciences fit requisites a, b, c---and thus can be the object of a hermeneutic science. (Query? Does meeting requirements a, b, c, entail that the hermeneutic method is the only way to study the human sciences?)

6. What is science of human behavior about? Behavior as action.

7. P. 189 "[A] good prima facie case ...that men and their actions are amenable to explanation of a hermeneutical kind. There is therefore some reason to raise the issue and challenge the epistemological orientation that would rule interpretation out of the sciences of man." Query? Who tries to rule it out? Logical empiricists? Hempel?

8. Extended example: Taylor's discussion of (behavioral) political science. Brute data? Voting behavior. Absence of need for interpretation. Use of questionnaires (an'keeto) to identify beliefs and evaluations of agents. Scientific activity consists in framing and testing correlations among acts, structures, institutions, procedures, and beliefs*, evaluations*, etc.

9. Does identification of social reality with brute data such as the above eliminates the need for interpretation?


Outline of 9th lecture

Topic: Continuation of C. Taylor's "Interpretation and the sciences of man"

1. Taylor's account of how a "behavioral" (or positive, or empiricist) political science avoids the problem of addressing the meaning of political behavior.

2. Question of whether separation of data from interpretations is possible. An example from anthropology, using a questionnaire to determine degree of acculturation. (D.H. Thomas, Figuring Anthropology). Also see Geertz on thick description.

3. Taylor's approach is different. He attacks the way behavioral political science reconstructs (or sets limits to) social reality. The problem, as he sees it, is that what we think of as the object of (political) beliefs, desires, and so forth have no reality in this system. Compare with early (Bridgman 1929) operational definitions of how much something weighs, how long it is, etc. In this system, abstract terms such as weight, length have no referent. In any concrete case, we can say what it means for something to weigh, for example, 60 kilograms, and that is all that is needed for science. Bridgman allowed "symbolic operations", for example, with pencil and paper, to count as operations as well as physical activities such as placing a meter stick along the edge of a table. Attempts to extend the program of operational definition to all mental concepts, especially "intelligence." Critics regard intelligence as something that might be measured, more or less satisfactorily, by a test, but refuse to identify intelligence with the score on an IQ test, no matter how good the test is. What is the underlying reality that the test measures?

4. Taylor's analogous question applied to political science concerns what underlies the application of voting language to marking a ballot slip, or in more general terms, applying the language of political participation with certain physical and symbolic activities.

5. Example: presence/absence of language of negotiation (See Smith, Agrarian Origins of Modern Japan). Societies differ not just in language, but in (interrelated) social practices. A difference between social reality and physical reality. The latter is not dependent on our having a language to describe it. (Some would disagree!) Mutual dependence of language and practices, constitutive rules, performative utterances, linguistic ways of marking whether one does or doesnıt engage in the practice, etc. But this doesnıt mean that social reality is only linguistic conventions (in the way that some formalists or conventionalists identify religion with "god talk").

6. Intersubjective meanings. These are the meanings that are constitutive of the social practices that make up social reality: Such practices embrace institutional practices (banking--writing checks, making deposits, etc.; religion--praying, offering a sacrifice; democratic government--casting a vote, running for office; Winch would call these meanings "rules"; Cognitive anthropologists, ethnomethodologists, linguistic anthropologists would also call them rules, but would place them "inside the heads" of individuals. C. Geertz, a symbolic anthropologist, would call them "culture" or "cultural norms." Geertz, like Taylor and Winch, says they are public in the sense that they do not exist only in the minds of agents; they are shared; they are accessible to investigation by observing behavior, including linguistic behavior. They differ from consensus, which refers to the convergence of beliefs of individual agents. Consensus (lack of consensus) presupposes the existence of intersubjective meanings (a common language rooted in social practices); otherwise there could be no agreement or disagreement. Against a background of intersubjective meanings, we can investigate brute data, just as we can speak of data within a Newtonian theory or data within an Aristotelian theory.

7. Common meanings involve recognition or consciousness of shared beliefs, aspirations, goals, values, and a common reference point for public life of a society. Examples: The determination of French-Canadian provinces (Quebec) to preserve a francophone state identity. The determination of a small town in Indiana to have a championship basketball team. Such common meanings are a basis for community life, but may not involve consensus about how the meaning is articulated. (Example--should an activist or pacifist approach be taken to achieve the shared goal?)

8. Taylor objects to behavioral political science because it excludes study of 6 & 7 above. Even if one allows that subjective beliefs can be "captured" in behavioral political science, 6 & 7 cannot. Thus, important differences among societies remain invisible and comparisons cannot be made. Inappropriate generalizations, etc.

9. Example of failure of behavioral political science to explain contemporary social disorder. Attempt to treat legitimacy of government and civilization of work as subjective beliefs. "Ideology". Point of examples: need to study intersubjective meanings, or retreat to "irrationality" of individuals as an explanation of social upheaval.

10. Is a social science that can deal with 6 & 7 necessarily hermeneutical? T's inadequate characterization of methods of science. Difficulty in attempts to understand the other applies to physical as well as human science. Old (refuted) views about why social sciences cannot predict + incommensurability. Taylor's conclusion: science of man cannot be value free, is founded on intuitions, cannot predict, requires freedom from illusion, and so forth.


Outline of 10th lecture

Topic: Rationality and cognitive relativism.

Some references: Rationality and Relativism, edited by Martin Hollis and Steven Lukes; Rationality, edited by B., Wilson; " Understanding a primitive society." P. Winch, Amer. Phil. Quarterly I, 1964; "Do Azande and Nuer use a nonstandard logic?" M. Salmon, Man, N.S. 13, 1978.

1. Problem: How do we understand beliefs that are (1) apparently irrational and (2) culturally accepted? Examples: (a)"Twins are birds." (b)"Some, but not all, Zande men are witches. Witchcraft is inherited through the male line. All Zande men are related through the male line." (c) "Mercury (the element) is a cure for syphilis because Mercury (the god) is god of the marketplace and syphilis is contracted in the marketplace." (d) "Men are red macaws." (e) Some complicated cosmologies of preliterate peoples. (f) The cosmology of Aristotle. (g) Some contemporary religious doctrines, e.g., There are three persons in one God.

2. What is at stake?

3. Solutions offered:

4. Features of rationality

Minimal criteria? Logical consistency: (1) Every sentence (proposition, statement) is either true or false. (2) No sentence is both. (Noncontradiction, excluded middle) (3) Rules of inference? Modus ponens (If p then q; p. Therefore q) Simplification: (p and q. Therefore p.) Basic arithmetic? Identity of indiscernibles? Rule of inductive generalization? (4) Sufficiently similar response to basic perceptual data? (5) Taking appropriate means to secure desired ends; transitivity of preferences.


Outline of Lecture 11

* Please note the following corrections to the Syllabus for the course: Lectures in week 8 will be on June 6 and 8; in week 9, on June 13 and 15; in week 10, on June 20 and 22. June 27 is the last class day. Term papers are due on that day. Suggested length of term papers is 10 to 15 pages.

Topic: Continuation of discussion of rationality and relativism, with special focus on "alien" beliefs of tribal and traditional societies as well as theories of premodern science.

1. Elaboration of Sperber's claim that not all belief is commitment to a proposition. His account of propositional and semi-propositional representations: A conceptual representation that succeeds in identifying one and only one proposition is a propositional representation. (Example: Salmon lives at 1064 Highmont Road in Pittsburgh.) A conceptual representation that fails to identify one and only one proposition is a semipropositional representation. (1. Salmon lives on Highmont Road. 2. People of different cultures live in different worlds.)

2. Elaboration of Sperber's account of representational beliefs (in contrast to factual beliefs). A subject holds a representational belief R if he/she holds a factual belief of the form "The proper interpretation of R is true." Examples: "R is a dogma in our Church." "Einstein convincingly argued for R." "My teacher say R is true." Contrast awareness of a (supposed) fact, in the case of factual beliefs, with awareness of commitment to a representation

Utility of semipropositional representations in science and everyday life. Avoidance of contradiction. Contrast between human knowledge processors and computers. Nuer belief in the truth of "Twins are birds" versus their acceptance of a metaphorical claim.

3. Features of rationality. See #4 in outline of 10th lecture.

4. Assumption of a "bridgehead" necessary for identifying (translating) alien beliefs and belief systems (M. Hollis). Commonly shared standards of truth and inference, and a commonly shared core of beliefs whose content or meaning is fixed by application of the standards.

5. Status of the assumption. Lukes says the assumption of a bridgehead is a priori, but the content of the bridgehead is empirically determined. Principle of charity (Davidson) Principle of humanity (Grandy).

6. Problems of interpretation, assuming questions about bridgehead are settled.

7. Lukes's answer to Winchıs form of relativism. Distinction between Rationality (I) and Rationality (II).

8. Explanations of why "alien" beliefs are held. Four types of belief: true, rational; false, rational; true, irrational; false irrational.


Outline of 12th Lecture

Topic: Functional Explanation

(a) Review of Hempel's Deductive-Nomological and Inductive-Statistical models of scientific explanation. Important points: Not all explanations are causal explanations. Empirical testability of the explanans. Predictive and retrodictive capability of laws.

(b) Functional analysis of x (recurrent activity of behavior pattern) exhibits the contribution of x to the development or preservation of the system in which it occurs.

"The heartbeat in vertebrates has the function of circulating blood through the organism." (Hempel)

Note that not all effects are functions.

(c) Examples of functional analysis in anthropology, psychology, sociology.

Malinowski: Magic and religion enhance mental attitudes, which have immense biological value.

Radcliffe-Brown: Totemic rites in Australian aboriginal societies have the function of maintaining the continuity of social life.

Freud: Compulsive behavior has the function of binding anxiety.

Merton: Function of Hopi ceremonial rain dances is to reinforce group identity.

Manifest (ostensible) and latent (social) functions of human behavior.

Functional analysis vs. functional explanation.

Explanation in terms of consequences or ends. Teleological explanations of purposive or goal-directed behavior. Aristotle?s "final causes".

Hempel's logical form of functional explanation:

(Time = t)

(a) At t1 s functions adequately in a setting of kind c (characterized by specific internal and external conditions).

(b) s functions adequately in a setting of kind c only if a certain necessary condition n is satisfied.

(c) If trait i were present in s, then, as an effect, condition n would be satisfied.

__________________________________

(d) Hence, at t1 trait i is present in s.

Problems: Fallacious form of argument (affirming the consequent). Repairs?

Functional indispensability for a given item.

Functional alternatives.

Difficulty of meeting empirical requirements for explanation

Heuristic value of functional "explanations"

E. Nagel's contributions. Teleology Revisited Columbia U. Press, 1979.

Others' insistence on the causal import of functional explanation. L. Wright Teleological Explanations U. of California Press, 1976. G.A. Cohen, see text.


Outline of Lecture 13

Topic: Continuation of discussion of Functional Explanation.

1. Disanalogies between biological systems and social systems--specification of limits of the system; specification of what constitutes "health of a system." Difficulties raised for empirical verification of premises of explanatory argument.

2. Functionalism versus functional explanation.

3. Conservative nature of functionalism. Political circumstances surrounding anthropological work of British social anthropologists. Political implications of their findings.

4. Materialist social science

categories
Technology
Forces of production
Economic structure
Relations of production
Politics, Culture
Superstructure
examples

Tools, Raw materials, Forms of Agriculture, Labor skills

Property ownership, Wage labor, Market system, Slavery
State, Kinship, Legal system, Religion, Police, Ideology, Political parties, Family

@

General approach: (1) All societies must have institutions through which the basic subsistence needs of the population are met. Technology and labor must be organized to produce food, shelter, etc. Institutions must exist for distribution of goods, and these must be stabilized by higher level institutions. Economic activity typically produces surplus, and a struggle ensues to control the surplus. (2) Various features of a given society can be understood in terms of their suitability to the working of the production system. (Functional explanations of production systems through which basic human needs are satisfied.) (Marx, 1846 The German Ideology; 1867 Capital).

5. Marxist seems committed to functional explanations, but not to conservatism of functionalist. Growth of human powers forces society to adjust--institutions serve this development; class struggles; revolutions, etc.

6. What is required for satisfactory functional explanation: Biological model: "Theory entails that plants and animals have the useful equipment they do because of its usefulness, and specifies in what manner the utility of a feature accounts for its existence. (Natural selection is the mechanism or microfoundation.)

7. To provide a satisfactory functional explanation of some societal feature, one must show (a) that the feature has the useful effect attributed to it; and (b) that the feature exists or persists ("is there") because it has that effect.

Example: (a) Joking relationship has the effect of reducing tension between antagonistic pairs.

(b) That effect is why the joking relationship persists in the society. (???)

8. Explanation not complete without an answer to (7b). Heuristic value? Cohen--seek various forms of elaboration--purposive activity on the part of agents (rational decision making); Darwinian (chance variation, scarcity, and selection all present); Lamarckian (alteration as a result of use, with no intention on the part of the agent to change); Self-deception (operates through mind of the agent, without full acknowledgement of the agent).

9. Ideologies.

10. Adaptation of economic structure to productive forces: class that best meets the demands of production tends to dominate (See 7a.) For 7b, There is a "general stake in stable and thriving production", so class that can deliver this draws allies from other classes (rational decision microfoundation operating here).

11. Conclusion: Marxism depends on functional explanations, but this fact does not diminish its revolutionary force. In addition, the functional explanations have heuristic value, and can sometimes be provided with the appropriate microfoundations to make them genuine causal explanations.


Outline of Lecture 14

Topic: Kincaid's defense of functional explanations in the social sciences.

1. Difference between identifying function of social institution (e.g., "The function of the joking relationship is avoidance of hostility between mother-in-law and son-in-law") and offering a functional explanation of that institution ("The joking relationship not only has the function described above, but it persists in the society precisely because it has that effect"). Kincaid: " assertions that practices exist in order to promote their effects, not just that practices have effects."

2. Dismissal of claim that the essence of functional explanation lies in negative feedback relationships (Faia) or maintaining homeostasis (Stincombe, N. Weiner).

3. Criticism of Cohen's consequence account. C: "We have a cause, increase of scale, and an effect, economies of scale. ... the cause occurred because of its propensity to have that effect; the increase in scale occurred because the industry was of a sort in which increases in scale yield economies." Kincaid says that Cohen's formulation requires only correlations, not causes. K. also says that consequence laws are not necessary, but C's account does not seem to commit him to the view that whenever a feature is useful it will come to exist.

4. Kincaid's understanding of what it means to give a functional explanation of A (i.e., A's persistence) with respect to its function B: (1) A causes B. (2) A persists because it causes B. (3) A is causally prior to B.

5. Comparison to account of Larry Wright.

6. Given the above (4) meaning of functional explanation, the problem in specific cases is to provide empirical support for (1)-(3). While this may be difficult, it is not a special problem for the social sciences, but instead involves the same difficulties that occur elsewhere in trying to establish causal regularities: show correlation, eliminate spurious causes

7. To establish that when A exists, it raises the probability that B exists--

B= x1 A + e1.

To establish that A persists because the above condition holds--

P = x2 E + e2 .

8. Direct tests possible in principle if requisite data are available. Suggestion: Time series data to answer questions about first appearance and/or persistence. Differential survival by traits, with mechanisms that link contributions to survival with persistence.

9. Indirect tests. Design analyses and optimality considerations.

10. Problems with all these possibilities.

11. Complexities in causal models. See p. 422.

12. Examples of failures.

13. Examples of successful functional explanations.


Outline of Lecture 15

Topic: Continuation of discussion of functional explanations. Empirical confirmation of the causal claims that constitute functional explanations. (A causes B. A persists because it causes B.)

Reference: Philosophical Foundations of the Social Sciences: Analyzing Controversies in Social Research, H. Kincaid, 1996, Cambridge University Press.

1. Indirect confirmation by showing that an existing practice is an optimal way to solve a problem in a given setting. Kincaid says (p. 421) "Harris's charming explanations of seemingly irrational taboos and practices likewise work by showing that the practice in question is really optimal, given the local environment.

Cultural Materialism: it is possible to fully explain human culture in economic, technological, and ecological terms.

C.M. analysis of Hindu's refusal to eat beef. "Beef aversion exists in order to ensure the economic survival of Indian farmers."

What is maximized (What does "economic survival" mean?)

Is practice optimal? Cost accounting.

What is the mechanism for selecting the optimal practice?

If practice is optimal, does it persist because it is optimal?

Harris's rejection of nonmaterial explanations for cultural practices.

Conclusion: Not very convincing case.

2. Evidence for functional explanations in Marxism. Vague claim; "The state exists in order to promote the interests of the ruling class." Evidence:

Criticisms: (a) Other theories (e.g. "politicians pursue their own self-interest") may predict this behavior as well as Marxism. (b) Origin and present function are different. (c) Even if true, M. must show that this causes persistence of the capitalist system. (d) Having ties is a symmetric relationship‹which way does the causal arrow point?

Basic problem: A's causing B can be spuriously correlated with A's persistence. (See figures in Kincaid's paper)

Alternate accounts: State pursues its own interests.

State sometimes supports working-class interests.

Conclusion: Evidence is inconclusive.

Kincaid: The failures of 1 and 2 are contingent and eliminable.

3. The ecology of organizations. Ecological models used to explain organizational change. Hannan and Freeman, 1989.

Goal of organizational ecology: to explain the kinds of organizations that exist, their relative numbers, and how those kinds of relative numbers change. An attempt to understand organizations at the population level. H & F contend that analysis at this level is required to explain large-scale social change.

Earlier models for studying organizations focused on individual, successful organizations and tried to extrapolate their characteristics. Problems with this type of analysis. Features of H and F's model:

Testing the model:

Conclusion: Excellent work, as well confirmed as good work in the nonexperimental natural sciences.

4. Final defense of functional explanations.

Confirmation possible without knowledge of underlying mechanisms.

Reliance on prerequisite analysis for optimality arguments.

Problems of evidence.

Failure to explain origins, illegitimate analogy with natural selection.


Outline of Lecture 16

Topic: Individualism vs. Holism; Durkheim on social facts.

1. Emile Durkheim (1858-1917): Some biographical details.

2. Important books:

3. Some key concepts: Alienation; Anomie; Functionalism

4. Biographies of Durkheim in English: Talcott Parsons's article in International Encyclopedia of the Social Sciences; A. Giddens's Emile Durkheim; S. Lukes's Emile Durkheim, His Life and Work: A historical and critical study.

5. Social facts provide the subject matter for a distinct (distinct from biology and psychology) science of sociology.

6. "ways of acting, thinking, and feeling" that are inherited through education, function independently of individual's use of them, exist outside individual consciousness, endowed with coercive power, exercise a check or constraining power on individual action, etc. Df. "ways of acting, thinking, and feeling, external to the individual, and endowed with a power of coercion, by reason of which they control him."

7. Reasons why social facts are neither biological nor psychological facts.

8. Problems about autonomy?

9. Examples of social facts (compare to Winchıs "rules") that exist where there is social organization.

10. Social currents, another kind of social fact. These "do not originate in any one of the particular individual consciousnesses." "collective manifestations" (Contrast individualists' account of "crowd psychology," but see Durkheim, p. 437.)

11. Durkheimıs argument for the coercive force of social facts.

12. Durkheimıs account of education: "a continuous effort to impose on the child ways of seeing, feeling, and acting, which he could not have arrived at spontaneously." Analog to "the socialization of the human being."

13. Universality (which does not define the social) versus collective aspects of the beliefs, tendencies and practices of a group.

14. Use of statistics to "isolate" social phenomena. Expression of "group mind."

15. Crux of debate between holists and individualists: " It [the social fact] is a group condition repeated in the individual because imposed on him. It is to be found in each part because it exists in the whole, rather than in the whole because it exists in the parts." (p.436)

16. Externality, coercion and generality: "If a mode of behavior whose existence is external to the individual consciousnesses becomes general, this can only be brought about by its being imposed on them." (Durkheimıs "coercion" vs. Tardeıs "imitation")

17. Domain of sociology: social facts. "A social fact is to be recognized by the power of external coercion which it exercises or is capable of exercising over individuals, and the presence of this power may be recognized in its turn either by the existence of some specific sanction or by the resistance offered against every individual effort that tends to violate it."

18. Social phenomena are things (data, objective facts, and the basis for a science).

19. Rule: The voluntary character of a practice or an institution should never be assumed beforehand.

20. Possibility of modifying social facts.

21. Comparison of reforms (from subjectivity to objectivity) in psychology and sociology. Embodiment of social facts in codes, statistical figures, costumes, architecture, art. Accessible, but complex and difficult to interpret.


Outline of Lecture 17

Topic: Holism vs. Individualism

Famous holists: Durkheim, Marx, Radcliffe-Brown, T. Parsons Contemporary holists: Kincaid, Lukes?, Mandelbaum, Goldstein.

Famous individualists: K. Popper, Contemporary individualists: Popperians at LSE (Watkins, R. Miller), D. Papineau, D. Little?

Attempts to make holist/individualist distinction less crude.

1. Methodological Individualism

Methodological Individualism (MI) denies the independent existence of things like the group mind, culture as a superorganic entity, and so forth.

Durkheim insists on the existence of social facts, but this could be consistent with the ontological thesis of MI. "Social norms are embodied in the behavioral dispositions of individuals," is consistent with Durkheimıs view that norms are external to the individual and carry coercive power.

1. Genetic makeup, brain states; 2. Aggressions, gratification; 3. Cooperation, power, esteem; 4. Cashing checks, saluting, voting. Attempts to define the social exclusively in terms of type 1 or type 2 predicates seem doomed to failure. Types 3 and 4 are social predicates. Quick conclusion: Social concepts cannot be eliminated by defining them in terms of concepts that have no social meaning. Would such elimination be desirable if it were possible? What role to concepts play in science? Answer: They are useful for analyzing and explaining the range of phenomena to which they apply.

2. Some issues to discuss:


Outline of Lecture 18

Topic: Objectivity in Science

1. "Science is concerned with fact, not values." "Any concern with values contaminates the objective character of science."

2. A rough sketch of the course of scientific research on a particular problem (from the scientistıs point of view)

3. Intellectual values vs. other sorts of value

4. Discussion of values associated with A- G above.

5. Values involved in scientific activity from the point of view of those dispensing resources:

6. Differences between natural sciences and physical sciences with respect to "intrusion" of values.


Outline of Lecture 19

Topic: Continuation of discussion of objectivity in science

1. Values involved in scientific activity from the point of view of those dispensing resources (see Outline for Lecture 18).

2. Differences between social and physical sciences with respect to "intrusion" of values.

3. Max Weber (1864-1920); biographical sketch.

4. Weber on objectivity: "There is no absolutely "objective" scientific analysis of . . . "social phenomena" independent of special . . . viewpoints according to which . . . they are selected, analyzed, and organized for expository purposes.

5. Quantitative vs. qualitative features of science.

6. Formulation of general concepts and discovery of laws as a goal of physical sciences.

7. Role of laws in social sciences as an aid to understanding significance of historical configurations.

8. Importance of causal knowledge in physical and social sciences.

9. Possibility of objective causal knowledge in the social sciences.

10. Rejection of notion of a "closed" (definitive, objectively valid, systematically fixed set of problems, etc.) social science.

11. Possibilities for objective studies of values:

12. Weber's claim that the problems of the empirical disciplines are not problems of evaluation and are to be solved "nonevaluatively."


Outline of Lecture 20

Topic: Critical Social Science. Reading: D. Comstock, "A Method for Crtitical Research"

1. Some differences between critical social science and "positive" social science (from the viewpoint of the Critical Social Scientist).

2. A method for critical social science; (Comstock, p. 632)

3. Features of critical social sciences: a method of praxis (combines disciplined analysis with practical action. Aims at changing the world, not merely understanding it,. Enables subjects to change their situation. Democratic rather than elitist.

4. What revisions, if any, should be made in Comstock's account of positive social science? (see Table 40.1, p. 632)

5. Are critical social science and positive social science (correctly interpreted) compatible with one another?

6. What is the relationship between critical social science and interpretive social science?

7. Is critical social science the only legitimate approach to social science? (i.e., the only approach that avoids manipulation, doesn't treat humans as objects rather than subjects, and so forth?)

Back to Philosophy of Social Science


Last modified Dec. 14, 2008. (c) Merrilee H. Salmon

@