From Robert K. Merton, Social Theory and Social Structure.Glencoe, IL: Free Press, 1957, pp. 60 – 69.
MANIFEST AND LATENT FUNCTIONS
As has been implied in earlier sections, the distinction betweenmanifest and latent functions was devised to preclude the inadvertentconfusion, often found in the sociological literature, betweenconscious motivations for social behavior and its objectiveconsequences. Our scrutiny of current vocabularies of functionalanalysis has shown how easily, and how unfortunately, the sociologistmay identify motives with functions. It was furtherindicated that the motive and the function vary independently andthat the failure to register this fact in an established terminologyhas contributed to the unwitting tendency among sociologists toconfuse the subjective categories of motivation with the objectivecategories of function. This, then, is the central purpose of oursuccumbing to the not-always-commendable practice of introducing newterms into the rapidly growing technical vocabulary of sociology, apractice regarded by many laymen as an affront to their intelligenceand an offense against common intelligibility.
As will be readily recognized, I have adapted the terms “manifest”and “latent” from their use in another context by Freud (althoughFrancis Bacon had long ago spoken of “latent process” and “latentconfiguration” in connection with processes which are below thethreshold of superficial observation).
The distinction itself has been repeatedly drawn by observers ofhuman behavior at irregular intervals over a span of many centuries.(64) Indeed, it would be disconcerting to find that a distinctionwhich we have come to regard as central to functional analysis hadnot been made by any of that numerous company who have in effectadopted a functional orientation. We need mention only a few of thosewho have, in recent decades, found it necessary to distinguish intheir specific interpretations of behavior between the end-in-viewand the functional consequences of action.
George H. Mead (65): “. . . that attitude of hostilitytoward the law breaker has the unique advantage [read: latentfunction] of uniting all members of the community in the emotionalsolidarity of aggression. While the most admirable of humanitarianefforts are sure to run counter to the individual interests of verymany in the community, or fail to touch the interest and imaginationof the multitude and to leave the community divided or indifferent,the cry of thief or murderer is attuned to profound complexes, lyingbelow the surface of competing individual efforts, and citizens whohave [been] separated by divergent interests stand together againstthe common enemy.
Emile Durkheim’s (66) similar analysis of the socialfunctions of punishment is also focused on its latent functions(consequences for the community) rather than confined to manifestfunctions (consequences for the criminal).
W. G. Sumner (67): “. . . from the first acts by whichmen try to satisfy needs, each act stands by itself, and looks nofurther than the immediate satisfaction. From recurrent needs arisehabits for the individual and customs for the group, but theseresults are consequences which were never conscious, and neverforeseen or intended. They are not noticed until they have longexisted, and it is still longer before they are appreciated.”Although this fails to locate the latent functions of standardizedsocial actions for a designated social structure, it plainly makesthe basic distinction between ends-in-view and objectiveconsequences.
R. M. MacIver (68): In addition to the direct effectsof institutions, “there are further effects by way of control whichlie outside the direct purposes of men . . . this type of reactiveform of control . . . may, though unintended, be of profound serviceto society.
W. I. Thomas and F. Znaniecki (69): ”Although all thenew [Polish peasant cooperative] institutions are thus formed withthe definite purpose of satisfying certain specific needs, theirsocial function is by no means limited to their explicit andconscious purpose . . . every one of these institutions–commune oragricultural circle, loan and savings bank, or theater–is not merelya mechanism for the management of certain values but also anassociation of people, each member of which is supposed toparticipate in the common activities as a living, concreteindividual. Whatever is the predominant, official common interestupon which the institution is founded, the association as a concretegroup of human personalities unofficially involves many otherinterests; the social contacts between its members are not limited totheir common pursuit, though the latter, of course, constitutes boththe main reason for which the association is formed and the mostpermanent bond which holds it together. Owing to this combination ofan abstract political, economic, or rather rational mechanism for thesatisfaction of specific needs with the concrete unity of a socialgroup, the new institution is also the best intermediary link betweenthe peasant primary-group and the secondary national system.”
These and numerous other sociological observers have, then, from timeto time distinguished between categories of subjective disposition(“needs, interests, purposes”) and categories of generallyunrecognized but objective functional consequences (“uniqueadvantages,” “never conscious” consequences, “unintended . . .service to society,” “function not limited to conscious and explicitpurpose”).
Since the occasion for making the distinction arises with greatfrequency, and since the purpose of a conceptual scheme is to directobservations toward salient elements of a situation and to preventthe inadvertent oversight of these elements, it would seemjustifiable to designate this distinction by an appropriate set ofterms. This is the rationale for the distinction between manifestfunctions and latent functions; the first referring to thoseobjective consequences for a specified unit (person, subgroup, socialor cultural system) which contribute to its adjustment or adaptationand were so intended; the second referring to unintended andunrecognized consequences of the same order.
There are some indications that the christening of thisdistinction may serve a heuristic purpose by becoming incorporatedinto an explicit conceptual apparatus, thus aiding both systematicobservation and later analysis. In recent years, for example, thedistinction between manifest and latent functions has been utilizedin analyses of racial intermarriage, (70) social stratification, (71)affective frustration, (72) Veblen’s sociological theories, (73)prevailing American orientations toward Russia, (74) propaganda as ameans of social control, (75) Malinowski’s anthropological theory,(76) Navajo witchcraft, (77) problems in the sociology of knowledge,(78) fashion, (79) the dynamics of personality, (80) nationalsecurity measures, (81) the internal social dynamics of bureaucracy,(82) and a great variety of other sociological problems.
The very diversity of these subject-matters suggests that thetheoretic distinction between manifest and latent functions is notbound up with a limited and particular range of human behavior. Butthere still remains the large task of ferreting out the specific usesto which this distinction can be put, and it is to this large taskthat we devote the remaining pages of this chapter.
Heuristic Purposes of the Distinction
Clarifies the analysis of seemingly irrational socialpatterns. In the first place, the distinction aids thesociological interpretation of many social practices which persisteven though their manifest purpose is clearly not achieved. Thetime-worn procedure in such instances has been for diverse,particularly lay, observers to refer to these practices as”superstitions,” “irrationalities,” “mere inertia of tradition,”etc. In other words, when group behavior does not– and,indeed, often cannot–attain its ostensible purpose there is aninclination to attribute its occurrence to lack of intelligence,sheer ignorance, survivals, or so-called inertia. Thus, the Hopiceremonials designed to produce abundant rainfall may be labelled asuperstitious practice of primitive folk and that is assumed toconclude the matter. It should be noted that this in no senseaccounts for the group behavior. It is simply a case of name-calling;it substitutes the epithet “superstition” for an analysis of theactual role of this behavior in the life of the group. Given theconcept of latent function, however, we are reminded that thisbehavior may perform a function for the group, although this functionmay be quite remote from the avowed purpose of the behavior.
The concept of latent function extends the observer’s attentionbeyond the question of whether or not the behavior attains its avowedpurpose. Temporarily ignoring these explicit purposes, it directsattention toward another range of consequences: those bearing,for example, upon the individual personalities of Hopi involved inthe ceremony and upon the persistence and continuity of the largergroup. Were one to confine himself to the problem of whether amanifest (purposed) function occurs, it becomes a problem, not forthe sociologist, but for the meteorologist. And to be sure, ourmeteorologists agree that the rain ceremonial does not produce rain;but this is hardly to the point. It is merely to say that theceremony does not have this technological use; that this purpose ofthe ceremony and its actual consequences do not coincide. But withthe concept of latent function, we continue our inquiry, examiningthe consequences of the ceremony not for the rain gods or formeteorological phenomena, but for the groups which conduct theceremony. And here it may be found, as many observers indicate, thatthe ceremonial does indeed have functions–but functions which arenon-purposed or latent.
Ceremonials may fulfill the latent function of reinforcing thegroup identity by providing a periodic occasion on which thescattered members of a group assemble to engage in a common activity.As Durkheim among others long since indicated, such ceremonials are ameans by which collective expression is afforded the sentimentswhich, in a further analysis, are found to be a basic source of groupunity. Through the systematic application of the concept of latentfunction, therefore, apparently irrational behavior may attimes be found to be positively functional for the group. Operatingwith the concept of latent function, we are not too quick to concludethat if an activity of a group does not achieve its nominal purpose,then its persistence can be described only as an instance of”inertia,” “survival,” or “manipulation by powerful subgroups in thesociety.”
In point of fact, some conception like that of latent function hasvery often, almost invariably, been employed by social scientistsobserving a standardized practice designed to achieve an objectivewhich one knows from accredited physical science cannot be thusachieved. This would plainly be the case, for example, withPueblo rituals dealing with rain or fertility. But with behaviorwhich is not directed toward a clearly unattainable objective,sociological observers are less likely to examine the collateral orlatent functions of the behavior.
Directs attention to theoretically fruitful fields ofinquiry. The distinction between manifest and latent functionsserves further to direct the attention of the sociologist toprecisely those realms of behavior, attitude and belief where he canmost fruitfully apply his special skills. For what is his task if heconfines himself to the study of manifest functions? He is thenconcerned very largely with determining whether a practice institutedfor a particular purpose does, in fact, achieve this purpose. He willthen inquire, for example, whether a new system of wage-paymentachieves its avowed purpose of reducing labor turnover or ofincreasing output. He will ask whether a propaganda campaign hasindeed gained its objective of increasing “willingness to fight” or”willingness to buy war bonds,” or “tolerance toward other ethnicgroups.” Now, these are important, and complex, types of inquiry.But, so long as sociologists confine themselves to the studyof manifest functions, their inquiry is set for them by practical menof affairs (whether a captain of industry, a trade union leader, or,conceivably, a Navaho chieftain, is for the moment immaterial),rather than by the theoretic problems which are at the core of thediscipline. By dealing primarily with the realm of manifestfunctions, with the key problem of whether deliberately institutedpractices or organizations succeed in achieving their objectives, thesociologist becomes converted into an industrious and skilledrecorder of the altogether familiar pattern of behavior. The termsof appraisal are fixed and limited by the question put to him by thenon-theoretic men of affairs, e.g., has the new wage-paymentprogram achieved such-and-such purposes?
But armed with the concept of latent function, the sociologistextends his inquiry in those very directions which promise most forthe theoretic development of the discipline. He examines the familiar(or planned) social practice to ascertain the latent, and hencegenerally unrecognized, functions (as well, of course, as themanifest functions). He considers for example, the consequences ofthe new wage plan for, say, the trade union in which the workers areorganized or the consequences of a propaganda program, not only forincreasing its avowed purpose of stirring up patriotic fervor, butalso for making large numbers of people reluctant to speak theirminds when they differ with official policies, etc. In short, it issuggested that the distinctive intellectual contributions ofthe sociologist are found primarily in the study of unintendedconsequences (among which are latent functions) of social practices,as well as in the study of anticipated consequences (among which aremanifest functions). (83)
There is some evidence that it is precisely at the point where theresearch attention of sociologists has shifted from the plane ofmanifest to the plane of latent functions that they have made theirdistinctive and major contributions. This can be extensivelydocumented but a few passing illustrations must suffice.
THE HAWTHORNE WESTERN ELECTRIC STUDIES: (84) As is well known, theearly stages of this inquiry were concerned with the problem of therelations of “illumination to efficiency” of industrial workers. Forsome two and a half years, attention was focused on problems such asthis: do variations in the intensity of lighting affect production?The initial results showed that within wide limits there was nouniform relation between illumination and output. Production outputincreased both in the experimental group where illuminationwas increased (or decreased) and in the control group where nochanges in illumination were introduced. In short, the investigatorsconfined themselves wholly to a search for the manifest functions.Lacking a concept of latent social function, no attention whateverwas initially paid to the social consequences of the experiment forrelations among members of the test and control groups or forrelations between workers and the test room authorities. In otherwords, the investigators lacked a sociological frame of reference andoperated merely as “engineers” (just as a group of meteorologistsmight have explored the “effects” upon rainfall of the Hopiceremonial).
Only after continued investigation, did it occur to the researchgroup to explore the consequences of the new “experimental situation”for the self-images and self-conceptions of the workers taking partin the experiment, for the interpersonal relations among members ofthe group, for the coherence and unity of the group. As Elton Mayoreports it, “the illumination fiasco had made them alert to the needthat very careful records should be kept of everything that happenedin the room in addition to the obvious engineering and industrialdevices. Their observations therefore included not only records ofindustrial and engineering changes but also records of physiologicalor medical changes, and, in a sense, of social andanthropological. This last took the form of a ‘log’ that gave as fullan account as possible of the actual events of every day. . . .” (85)In short, it was only after a long series of experiments which whollyneglected the latent social functions of the experiment (as acontrived social situation) that this distinctly sociologicalframework was introduced. “With this realization,” the authors write,”the inquiry changed its character. No longer were the investigatorsinterested in testing for the effects of single variables. In theplace of a controlled experiment, they substituted the notion of asocial situation which needed to be described and understood as asystem of interdependent elements.” Thereafter, as is now widelyknown, inquiry was directed very largely toward ferreting out thelatent functions of standardized practices among the workers, ofinformal organization developing among workers, of workers’ gamesinstituted by “wise administrators,” of large programs of workercounselling and interviewing, etc. The new conceptual scheme entirelyaltered the range and types of data gathered in the ensuing research.
One has only to return to the previously quoted excerpt fromThomas and Znaniecki in their classical work of some thirty yearsago, to recognize the correctness of Shils’ remark:
. . . indeed the history of the study of primarygroups in American sociology is a supreme instance of thediscontinuities of the development of this discipline: aproblem is stressed by one who is an acknowledged founder of thediscipline, the problem is left unstudied, then, some years later, itis taken up with enthusiasm as if no one had ever thought of itbefore. (86)
For Thomas and Znaniecki had repeatedly emphasized thesociological view that, whatever its major purpose, “the associationas a concrete group of human personalities unofficially involves manyother interests; the social contacts between its members are notlimited to their common pursuit. . . .” In effect, then, it had takenyears of experimentation to turn the attention of the WesternElectric research team to the latent social functions of primarygroups emerging in industrial organizations. It should be made clearthat this case is not cited here as an instance of defectiveexperimental design; that is not our immediate concern. It isconsidered only as an illustration of the pertinence for sociologicalinquiry of the concept of latent function, and the associatedconcepts of functional analysis. It illustrates how the inclusion ofthis concept (whether the term is used or not is inconsequential) cansensitize sociological investigators to a range of significant socialvariables which are otherwise easily overlooked. The explicitticketing of the concept may perhaps lessen the frequency of suchoccasions of discontinuity in future sociological research.
The discovery of latent functions represents significantincrements in sociological knowledge. There is another respect inwhich inquiry into latent functions represents a distinctivecontribution of the social scientist. It is precisely the latentfunctions of a practice or belief which are not common knowledge, forthese are unintended and generally unrecognized social andpsychological consequences. As a result, findings concerning latentfunctions represent a greater increment in knowledge than findingsconcerning manifest functions. They represent, also, greaterdepartures from “common-sense” knowledge about social life. Inasmuchas the latent functions depart, more or less, from the avowedmanifest functions, the research which uncovers latent functions veryoften produces “paradoxical” results. The seeming paradox arises fromthe sharp modification of a familiar popular preconception whichregards a standardized practice or belief only in terms of itsmanifest functions by indicating some of its subsidiary or collaterallatent functions. The introduction of the concept of latent functionin social research leads to conclusions which show that “social lifeis not as simple as it first seems.” For as long as people confinethemselves to certain consequences (e.g. manifestconsequences), it is comparatively simple for them to pass moraljudgments upon the practice or belief in question. Moral evaluations,generally based on these manifest consequences, tend to be polarizedin terms of black or white. But the perception of further (latent)consequences often complicates the picture. Problems of moralevaluation (which are not our immediate concern) and problems ofsocial engineering (which are our concern (87)) both take on theadditional complexities usually involved in responsible socialdecisions.
An example of inquiry which implicitly uses the notion of latentfunction will illustrate the sense in which “paradox”–discrepancybetween the apparent, merely manifest, function and the actual, whichalso includes latent functions tends to occur as a result ofincluding this concept. Thus, to revert to Veblen’s well-knownanalysis of conspicuous consumption, it is no accident that he hasbeen recognized as a social analyst gifted with an eye for theparadoxical, the ironic, the satiric. For these are frequent, if notinevitable, outcomes of applying the concept of latent function (orits equivalent).
64. References to some of the more significant among these earlierappearances of the distinction will be found in Merton,”Unanticipated consequences . . . ,” op. cit.
65. George H. Mead, “The psychology of punitive justice,” AmericanJournal of Sociology, 1918, 23, 577-602, esp. 591.
66. As suggested earlier in this chapter, Durkheim adopted afunctional orientation throughout his work, and he operates, albeitoften without explicit notice, with concepts equivalent to that oflatent function in all of his researches. The reference in the textat this point is to his “Deux lois de l’evolution penule,” L’anneesociologique, 1899-1900, 4, 55-95, as well as to his Divisionof Labor in Society (Glencoe, Illinois: The Free Press, 1947 ) .
67. This one of his many such observations is of course from W.G.Sumner’s Folkways, (Boston Ginn & Co., 1906), 3. Hiscollaborator, Albert G. Keller retained the distinction in his own.writings; see, for example, his Social Evolution, ( New York:MacMillan, 1927), at 93-45.
68. This is advisedly drawn from one of MacIver’s earlier works,Community, (London: MacMillan, 1915). The distinction takes ongreater importance in his later writings, becoming a major element inhis Social Causation, (Boston: Ginn & Co., 1942 ), esp. at.314-321, and informs the greater part of his The More Perfect Union,(New York: MacMillan, 1918).
69. The single excerpt quoted in the text is one of scores whichhave led to The Polish Peasant in Europe and America beingdeservedly described as a “sociological classic. See pages 1426-7 and1523 ff. As will be noted later in this chapter, the insights andconceptual distinctions contained in this one passage, and there aremany others like it in point of richness of content, were forgottenor never noticed by those industrial sociologists who recently cameto develop the notion of ”informal organization” in industry.
70. Merton, “Intermarriage and the social structure.” op.cit.
71. Kingsley Davis, “A conceptual analysis of stratification,”American Sociological Review, 1942, 7, 309-321.
72. Thorner, op. cit., esp. at 16 t.
73. A.K. Davis, Thorstein Veblen’s Social Theory, HarvardPh.D. dissertation, 1941 and “Veblen on the decline of the ProtestantEthic,” Social Forces, 1944, 22, 282-86; Louis Schneider, TheFreudian Psychology and Veblen’s Social Theory, (New York: King’sCrown Press, 1948), esp. Chapter 2.
74. A.K. Davis, “Some sources of American hostility to Russia,”American Journal of Sociology, 1947, 53, 174-183.
75. Talcott Parsons, “Propaganda and social control,” in hisEssays in Sociological Theory.
76. Clyde Kluckhohn, “Bronislaw Malinowski, 1884-1942,” Journalof American Folklore, 1943, 56, 208-219.
77. Clyde Kluckhohn, Navaho Witchcraft, op. cit., esp. at46-47 and ff.
78. Merton, Chapter XII of this volume.
79. Bernard Barber and L. S. Lobel, “‘Fashion’ in women’s clothesand the American social system, Social Forces, 1952, 31,121-131.
80. O. H. Mowrer and C. Kluckhohn, “Dynamic theory ofpersonality,” in J.M. Hunt, ed., Personality and the BehaviorDisorders, ( Nev York: Ronald Press, 1944), 1, 69-135, esp. at72.
81. Marie Jahoda and S. W. Cook, “Security measures and freedom ofthought: an exploratory study of the impact of loyalty and securityprograms,” Yale Law Journal, 1952, 61, 296-333.
82. Philip Selznick, TVA and the Grass Roots (University ofCalifornia Press, 1949); A. W. Gouldner, Patterns of IndustrialBureaucracy (Glencoe, Illinois: The Free Press, 1954), P. M.Blau, The Dynamics of Bureaucracy (University of ChicagoPress, 1955 ); A. K. Davis, “Bureaucratic patterns in Navy officercorps,” Social Forces 1948, 27, 142-153.
83. For a brief illustration of this general proposition, seeRobert K. Merton, Marjorie Fiske and Alberta Curtis, MassPersuasion, ( New York: Harper, 1946) 185-189; Jahoda and Cook,op. cit.
84. This is cited as a case study of how an elaborate researchwas wholly changed in theoretic orientation and in the character ofits research findings by the introduction of a concept approximatingthe concept of latent function. Selection of the case for thispurpose does not, of course imply full acceptance of theinterpretations which the authors give their findings. Amongthe several volumes reporting the Western Electric research, seeparticularly W. J. Roethlisberger and W. J. Dickson, Managementand the Worker, (Harvard University Press, 1939).
85. Elton Mayo, The Social Problems of an IndustrialCivilization, (Harvard University Press, 1945), 70.
86. Edward Shils, The Present State of American Sociology,(Glencoe, Illinois. The Free Press, 1948), 42 [italics supplied].
87. This is not to deny that social engineering has direct moralimplications or that technique and morality are inescapablyintertwined, but I do not intend to deal with this range of problemsin the present chapter. For some discussion of these problems seechapters VI, XV and XVII; also Merton, Fiske and Curtis, MassPersuasion, chapter 7.