Edited by Martin and McIntyre. 2001. Readings in the Philosophy of Social Science. Massachusetts Institute of Technology.
22: The Logic of Functional Analysis, by Carl G Hempel. Pages. 349-375.
23: Function and Cause, by R.P. Dore. P. 377-389.
24: Functional Explanation: In Marxism, by G. A. Cohen. P. 392-402.
25: Functional Explanation: In Social Science, by Jon Elster. P. 404-414.
26: Assessing Functional Explanations in the Social Sciences, by Harold Kincaid. P. 415-428.
I offer many quotes to describe functions as generally delineated to a political system. In doing so, I will help explain what is meant by the concept of a system; divulging traits of inputs and outputs, system integration, and system maintenance. I will also briefly articulate functional analysis as it relates to equilibrium, an explanation of social conflict, political change, and significant phenomena of change. First, I must explore the concept of functional analysis.
A definition of functional analysis varies from philosopher to philosopher. Hempel thought functional analysis enabled researchers to “understand a behavior pattern or a sociocultural institution in terms of the role it plays in keeping the given system in proper working order and thus maintaining it…” (Readings in the Philosophy of Social Science (R), 354, italics added). Dore believed it to be methodological and philosophical positions utilized to explain the functions and causes of an institution, and the “recurring particular events” within the system (R, 386).
Elster found that a major assumption within functional analysis was that it “can succeed only if there are reasons for believing in a feedback loop from the consequence to the phenomena” (R, 407). Malinowski’s Principle denoted that “all social phenomena have beneficial consequences that explain them” (R, 404). Finally, Kincaid broadly defined functional analysis to be that which “explains social phenomena by means of their functions” (R, 415).
The “system” is the thing to which inputs and outputs may be observed as functional or dysfunctional. According to Hempel, the “system” is the observed conception, “tacit or expressed,” within functional analysis (R, 365). The bureaucratic system, for example, is observable and contains many manifest and latent functions. Dore, for example, found that social systems are “mediated by the personality structure… take a long time interval to work through the whole causal sequence” (R, 382). Self-regulating systems are causal because “initial states” produce “the same kind of final state” (R, 369). Kincaid determined that something could be functionally explained without being part of a system. Kincaid wrote that the best way to confirm functional account is to “confirm that A exists in order to fulfill its function B” (R, 419). For example, the bureaucracy exists in order to implement legislation.
Inputs and outputs are the variables within functional analyses (that often reveal a causal nature). The Markov-chain theory, for example, would stop changing a system once a “Yes” input was found (R, 408). Natural selection, on the other hand, takes inputs and maximizes the evolutionary output, regardless of the temporal variable. Likewise, Elster found that “bureaucracies sometimes survive merely by doing no harm” (Ibid). The lack of obvious dysfunctions provided a stable equilibrium. Overall, inputs and outputs may be seen as the “traits… effects… properties” within the functional explanation (R, 424).
System integration and system analysis allows researchers to comprehend “self-regulatory phenomena… in terms of purposes and motives” (R, 370). Manifest functions are the primary system purposes, while latent functions are unintended manifestations within the system (R, 379). Researchers are not expected to describe all aspects of system integration, but this analysis does allow for a determination of “function to cause” (R, 380). A change within the system integration may have violent effects for the overall system and surrounding systems. For example, if society changes the bureaucratic system “from R to F, or from F to R, all other subsystems may be thrown out—or move away from—equilibrium” (R, 409).
Functional analysis, in general, states that each system has some type of normal equilibrium. The system is self regulating because dysfunctions are removed and the functional system returns to equilibrium. Serious dysfunctions may destroy the system and a new system of some manner will form. Hempel urged that every part of a system “exists as a means to an end” (R, 371). The bureaucratic system, for example, might reveal latent manifestations which cause dysfunction, but may be remedied in order for a return to equilibrium. Here, the effect is “maintenance of ‘social equilibrium’ or societal survival” (R, 417).
Political change occurs in the political system, according to functional analysis, when the system is dysfunctional. According to Merton, dysfunctions are “observed consequences which lessen the adaptation or adjustment of the system” (R, 366). Once the system begins to reveal dysfunctions, a “strong and insistent” pressure to expunge the dysfunctions will surface and execute (R, 367). The system is hence “self-regulating” and equilibrium is witnessed once dysfunctions are unknown (R, 368). However, Elster noted that the integration of the political system may permit or deny conflict and/or adaptation.
Political change may not occur very often, since the political system constantly self-regulates in order to maintain equilibrium. Kincaid found that “surviving organizations rarely change their basic characteristics. Those traits thus exist because of their effects” (R, 424). Once the researcher is familiar with the inputs and outputs within a system, they may be able to predict social conflict. Kincaid recommended that researchers making predictions should evaluate “what an entity must be like to survive and then comparing that prediction with actual cases” (R, 421). The comparative political scientist studying efficient bureaucracies; for example, should study the variables of similar bureaucracies. More precisely, the researcher might recommend efficiency solutions—based on findings from similar bureaucracies determined to perform at greater or lesser efficiency.
Significant phenomena of change originate in various manners. According to Hempel, phenomena of change depended on the facts and laws “specified in the explanans” (R, 352). He deduced precisely, “explanans contain general laws, it permits predictions concerning occurrences other than that referred to in the explanandum” (R, 353). Here, the scientist may empirically test the explanans (Ibid). Dore, likewise, realized that phenomena of change may advent as latent and manifest functions and cause a “self-fulfilling prophecy” (R, 379). Here, the equilibrium within the political system may create new phenomena which increases stability of the system—based on the input-output manifestations.
In conclusion, functional analysis has had a prominent position in political science, particularly in analyses of the political system. I have articulated functions as generally delineated to a political system. I have briefly explained what is meant by the concept of a system through an analysis of inputs and outputs, system integration, and system maintenance. I also articulated functional analysis as it relates to equilibrium, an explanation of social conflict, political change, and significant phenomena of change.