Response to Social Crisis and Disaster

Response to Social Crisis and DisasterAuthor(s): E. L. Quarantelli and Russell R. DynesReviewed work(s):Source: Annual Review of Sociology, Vol. 3 (1977), pp. 23-49Published by: Annual ReviewsStable URL: http://www.jstor.org/stable/2945929 .

Accessed: 01/02/2013 02:05

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review ofSociology.

http://www.jstor.org

This content downloaded on Fri, 1 Feb 2013 02:05:12 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=annrevs

http://www.jstor.org/stable/2945929?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Ann. Rev. Sociol. 1977. 3.23-49 Copyright i) 1977 by Annual Reviews Inc. All rights reserved

RESPONSE TO SOCIAL CRISIS *.10533

AND DISASTER

E. L. Quarantelli and Russell R. Dynes Disaster Research Center, Department of Sociology, Ohio State University, Columbus, Ohio 43201

INTRODUCTION

Three decades ago, there was not enough theoretical material or research work on response to social crisis and disaster to have warranted writing more than a footnote attesting to that fact. A paragraph could have summarized all the relevant work and activity of two decades ago. Even ten years ago, several pages might have sufficed to summarize totally the burgeoning activities in the area. The growth of the area has been exponential, not linear, and today the research efforts and theoretical advances have reached such a level as to force us to be highly selective in our coverage in this paper.

The very concept of crisis has been reformulated and refined into and under a more generic category, collective stress situation. Barton (1970:38) defines such a situation as one where ". . . many members of a social system fail to receive expected conditions of life from the system." In turn, the notion of consensus and dissensus types of crisis has been advanced and fruitfully applied in research efforts (Warheit 1968, Waxman 1973a). Dissensus types of crises are conflict-containing situations where there are sharply contrasting views of the nature of the situation, what brought it about, and what should be done to resolve it (Stallings 1973). The campus civil disturbances of the late 1960s in the United States and western Europe illustrate this type of crisis well. Consensus-type crises are those where there is agreement on the meaning of the situation, the norms and values that are appropriate, and priorities that should be followed (Quarantelli 1970). Natural disasters, as well as those disasters occasioned by technological agents, are major examples of this type of crisis. The focus of this paper is almost exclusively on the research and theory now extant about mass and group response to disasters. While the recent work on riots is useful comparative material, it mainly highlights the fact that disasters are a different kind of collective stress situation. Looting, for example, has been frequent in recent American urban riots (Janowitz 1968), whereas it is generally absent in natural disasters in the Western world (Quarantelli & Dynes 1970c).

23



24 QUARANTELLI & DYNES

The term disaster itself has undergone a number of efforts at reformulation, with some more successful than others. At least four major references for the term have been noted: the physical agent, the physical consequences of the agent, the way in which the impact of the physical agent is evaluated, and the social disruption and social changes brought about by the physical agent and its impact (Dynes 1974). Almost all recent definitions use some version of this last conception, following the first socially oriented definition innovatively advanced by Fritz (1961). In this respect, definitions of a social nature have clearly and fortunately replaced the very early referents in almost solely physical terms. Nevertheless, even the newer concep- tions tend to assume concentrated space-time events, leaving unclear the categorical status of very diffuse events, such as famines and epidemics, that would otherwise be classified as disasters. In fact, some writers have stated that the emphasis on a specific event as an identifying feature is a pro-Western, pro-technology, pro-capital- ism bias, unsuitable for distinguishing disasters in underdeveloped societies (West- gate & O'Keefe 1976). Other critics have argued that disasters are inherently political phenomena and should be so conceptualized (Brown & Goldin 1973); the implication of this for research, if it is a valid position, has so far been unrecognized. The most extreme critical attack is that the word is an outmoded concept, a residue from the sweep of history that captures relatively insignificant phenomena instead of the newer terrors that have emerged in the modern world (Barkun 1974). All of these critics of the concept seem to be making some valid points, but none offers a satisfactory reconceptualization or new terms useful for analytical and research purposes. A few writers have tended to use the term natural hazard in place of disaster, but this seems a regression to earlier physical rather than social referents and totally ignores catastrophes generated by technological rather than natural agents.

Therefore, in this paper we selectively summarize and highlight basic substantive and structural trends only from the literature that either implicitly or explicitly assumes that a disaster is primarily a social phenomenon and is thus identifiable in social terms. We also restrict ourselves to reacting to research that aims at general propositions about social responses to actual disasters or to general discussions intended to contribute to theoretical advances, rather than addressing operational, planning, or administrative problems. Consequently, for purposes of this review we do not consider such work as sheer historical or geographic descriptive accounts, like those that characterize much of the work under the label of "natural hazards;" the studies, often of a laboratory or experimental nature, involving stressful but nondisaster contexts, as is true of much of the research on risk taking; or attitudinal surveys of hypothetical and potential, but seldom actual, disaster situations, as is the case of many studies in the pollution area. In this respect, the corpus of material now available on disaster-related phenomena is far greater than the sociologically oriented sources to which we allude, which are themselves of fairly hefty bulk.

In our original commitment to writing this paper, we assumed we might present a codification of research findings organized in a matrix that would cross-classify major disaster tasks with different levels of analysis. However, as we soon illustrate, since then several major works on general codification, as well as a series of codifica-



SOCIAL CRISIS AND DISASTER 25

tion attempts on more specific topics, have appeared or are under way. Therefore, to avoid unnecessary overlap we have decided instead to point out what has so far been ignored in the disaster literature. This includes some major changes and alterations in trends, of both a substantive and structural nature, in the theories and research prevailing in recent sociological studies of disasters. For our purposes, "recent" refers to the last ten or fifteen years, since pre- 1960 materials and activities are so well summarized elsewhere (Fritz 1961, 1968).

Among recent major substantive trends in the disaster field have been the following: both specific and general codification efforts; an even greater social organizational rather than social psychological emphasis; major attempts to use groups rather than individuals as the basic unit of study; an increasing employment of "system" notions in field research and analyses; attempts to merge collective-behavior and complex-organization approaches in studies of organized behavior in disasters; an accelerating attention to factors in the pre-impact period of disasters as determinants of trans- and post-disaster responses; a developing focus on long-run functional as well as dysfunctional aspects of disasters; and the initiation of model building. Among major structural trends have been the institutionalization of disaster research in academic settings, the perceived relevance of disaster research to public policy and agency responsibility, and the development of a "critical mass" of social and behavioral scientists interested in disaster research. We do not consider all these changes in substance and structure to be necessarily totally positive, nor do we feel that they address all relevant matters in an emerging field of study. We do believe that these changes have set the stage for what appears to be the start of a highly productive leap in research and theory in the next few decades in the disaster area. In fact, we conclude this paper by making explicit some of the many gaps and challenges that future disaster researchers ought to tackle.

SUBSTANTIVE TRENDS

Efforts at Codification If codification of research findings is one indicator that an area of inquiry is matur- ing, the recent sociological work on disasters marks the coming of age of the area. There have been several major attempts at general codification, efforts at overall systematization of the findings on a variety of questions from diverse studies. We find even more numerous but limited endeavors to collate and to order what is known about specific disaster-related processes, e.g. the topic of warning, or particular disaster problems such as the topic of antisocial behavior.

The first general codification of findings was done by Fritz (1961) in the infancy of the history of disaster studies. Despite the initial obscurity of his article, as a result of its placement in a general social-problems text, it eventually became widely known to disaster students, was and is often cited, and has attained the status of a minor classic work in the area. More than 15 years later, many of the significant research questions and issues raised by this article have not been followed through. For example, almost no systematic work has been done to examine what Fritz indicated that earlier accounts suggested about the positive or functional aspects of




disasters, the importance of the "therapeutic community" in softening the psychological impact for victims, or the "amplifier rebound" effect in disasters, which carries the stricken locality beyond its pre-impact levels of integration, productivity, and capacity for growth.

A decade after the Fritz article, Barton (1970) wrote what is probably the best known and the most sophisticated of the general codification efforts. In a paperback edition that attained wide circulation, Barton singles out, in sociologically relevant language, key problems of individual behaviors in disasters, such as the matter of role definition, role competence, and possible role conflict during such emergencies. A network of hypotheses are advanced, linking motivation, knowledge of what to do, and existence of social relationships with the probable disaster role that is likely to ensue, given combinations of the three indicated factors. Another major section of the book addresses, again in sociological terms, the question of the coordination of individual and organizational behaviors in disasters. Through a series of dia- grams, the complex interrelationships between such factors as organizational mobilization and the rate of nonadaptive behavior of individuals in the mass assault upon the emergency are highlighted. The climax of the book, however, is in the setting forth of a very intricate theoretical model to explain and to predict the rise of the "therapeutic community" and what factors affect individual behavior with respect to that community. A total of 71 different propositions are advanced, 39 of them referring to individual-level relationships, 23 to contextual relationships, and 9 to collective-level relationships. The propositions range from the simple and fairly obvious to the complex and the less apparent. For example, proposition # 1 is "The greater the severity of the impact the greater the number of victims" (1970:218); proposition #45 is "Subjective deprivation makes for less sympathy and identifica- tion with victims; conversely, the more people feel relatively better off than others, the more concerned they are with the deprived." (1970:245).

Barton's treatise would be an impressive achievement in almost any area. He included research studies and accounts by a variety of different researchers with varying competencies, using various methodologies and operating out of a range of disciplinary backgrounds. Yet he was able to recast many of the observations and findings in standard sociological language, to discern common elements in the heterogeneous observations, and to derive networks of empirically testable hypotheses as well as several overall theoretical frameworks within which the sets of hypotheses could be located. Nevertheless, as Barton himself admits, almost none of the propositions rest on solid empirical data; for many, only illustrative examples can be cited for support in the research literature. In many ways the Barton work was an exercise-an excellent one, to be sure-in hypothesis formulation and theory generation, more than it was the systematization of empirical generalizations. Nev- ertheless, it stands as the most genuinely sociological of all the general works on disasters written to the present time and offers to any enterprising researcher a multitude of interrelated hypotheses that could be put to an empirical test. While Barton himself favors quantitative survey-type studies, not all that he advances for research possibility would require that particular kind of research technique. For some of his more holistic propositions, for example, negative case studies could serve the appropriate research purposes just as well as population surveys.




Another recent codification effect was undertaken by Dynes in Organized Behav- ior in Disaster (1974; originally published in 1970). This book does not, as has sometimes been mistakenly assumed, attempt to pull together the research findings of the Disaster Research Center (DRC), which Dynes helped to establish in 1963; at the most, only illustrative material is used from the DRC work. The bulk of the book draws from about 250 pre-DRC accounts about organized behavior in disasters and especially focuses on the literature that either descriptively or analytically reports on the impact of disasters on formal or complex organizations in American society. A basic idea advanced is that organizational mobilization and recruitment of personnel, and the operational problems of adapting to radically changed environ- mental conditions, can be examined best by separating out four different types of groups likely to respond to disasters: established, expanding, extending, and emergent organizations. An attempt is also made to show how interorganizational relationships are affected by boundary personnel, organizational sets, and organizational legitimacy, and how a community disaster structure emerges from the creation and coordination of task subsystems. Little attention is paid, however, to findings on individuals and families, or at the other extreme, about societies or higher-level units.

A very recently published book, Human Systems in Extreme Environments: A Sociological Perspective (Mileti et al 1975), opens with the statement that "it is an attempt to codify what the social science literature reveals about how humans, individually and collectively adapt and respond to natural hazards and disasters" (1975:iii) As was done earlier by Fritz, Barton, and Dynes, findings were extracted across the entire spectrum of disaster-related activities, such as preparedness, warning, impact, the post-impact period, etc. The research literature review, as was true also in the instance of the just-mentioned codifiers, was selective-in this case it was limited to mostly published material, which totaled 191 items in all, although it included very few of the current specific efforts at codification, which we discuss later. Unlike the earlier general codification attempts, in this effort findings were systematically extracted across a range of levels, from individual to group, organization, community, society, and to the international level. This latter format allowed the production of a quantitative depiction of the uneveness of disaster research at different levels of study. Thus, the authors report "that the largest segment of findings was at the level of the individual; the community level was second" (1975:9). Furthermore, the findings were coded into a 36-cell "knowledge matrix" that was developed from cross-tabulating the variables of system level and time. The authors found a total of 1399 variables that could be placed into the matrix.

However, the authors self-consciously made no attempt to derive new propositions from the works they examined or to fit the findings into some kind of integrated theoretical scheme. In this sense, what results is a rather impressive, straightforward inventory (although it lacks a theoretical web) as was spun by Barton, and fails to uncover as many latent patterns and relationships as were surfaced by Dynes. However, although in this respect the work is less of a codifying effort than the others mentioned, it is a far more systematic inventory along other lines, and is the only existing work that tries to depict in quantitative terms the degree of current knowledge about different aspects of human and social behavior in disasters.




Other writers have selected out particular topics for codification. The works here range from somewhat general essays alluding to the literature reviewed to systematic listings of propositions keyed back to the particular referent source. An example of the latter is a recent monograph on Organizational Communications and Decision Making in Crises (Dynes & Quarantelli 1976). This review of only 35 sources, all but a few specifically dealing with disasters, generated 294 propositions. Of the total number, 126 treated decision making as the dependent variable and 21 treated it as the independent variable, while 101 treated communication as the dependent variable and 46 treated it as the independent variable. An even more elaborate kind of codification of a more advanced order was recently undertaken by Stallings (1977). Taking a long-used fourfold typology of organized behavior in disasters-established, expanding, extending, and emergent groups-he examined three empirical research efforts to see what major propositions the typology had generated. He was able to find empirical support for 18 organization-level propositions (e.g. extending organizations are seldom effectively controlled by other crisis-relevant organizations) and for 9 intraorganizational-level propositions (e.g. the greater the degree of internal autonomy possessed by subunits in established organizations, the more likely that tasks will be altered before structure in a disaster). He then links the typology of organized behavior in disasters to concepts and propositions derived from organizational theory in sociology. Seventeen propositions were generated that are new in the sense that they are not derived directly from disaster research as such; nonetheless, they could be tested in a disaster context involving the four types of groups (e.g. the greatest quantity of codified tasks-those requiring specialized expertise-will be found in established organizations, while the greatest number of generalized tasks will be found in emergent groups; expanding and extending organizations will occupy an intermediate position).

While all specific codification attempts are not as elaborate or as complex as the two just discussed, their very number indicates the high priority that such activity has recently acquired among disaster researchers. Among the topics that have been the subject of more or less codification efforts have been military-civilian relations during disaster operations (Anderson 1968), functioning of expanding organizations in community disasters (Dynes 1968), community functions under disaster conditions (Wenger & Parr 1969), police departments in disaster operations (Kennedy et al 1969), local civil defense in natural disasters (Anderson 1969b), warning systems in disaster situations (McLuckie 1970), communications in disasters (Stallings 1971), community conflict in natural disasters (Quarantelli & Dynes 1976), evacuation (Hultaker 1976), and panic behavior (Quarantelli 1977). That the codification fervor has perhaps not yet hit its peak may be indicated by the fact that we currently know of systematic work proceeding on codifying research findings with respect to a number of topics. At DRC alone, for example, such work is being pursued on looting and antisocial behavior, evacuation and evacuees, emergency medical service delivery, and emergency mental health services.

Particularly troublesome in all the codification efforts, general and specific, is a point very well documented by Mileti, Drabek & Haas (1975:9). They estimate that more than 95% of the findings they located were based on data taken from a single




set of events. While this may be a rather high figure for all codification efforts, it would appear, nevertheless, that many generalizations are drawn from one or a very few disaster studies. Furthermore, there is a tendency in the codification works to treat all data sources as having equal validity and acceptability. Propositions based on solidly derived evidence are sometimes given no more weight than statements based on chance observations. For example, some findings are arrived at through very sophisticated and sometimes longitudinal field studies using appropriate sam- ples, as in the case of such researchers as Drabek and his colleagues (1973, 1975, 1976) examining long-run consequences of disasters on families. Yet such findings are, on occasion, treated as equivalent to journalistic-like impressions garnered by sociologists who may visit disaster areas but who, with respect to sampling, prepara- tion of questions, systematic analyses of data, etc, simply do not carry out anything that in most interpretations would qualify as a sociological field study.

On the other hand, more encouraging is the fact that the generalizations being advanced by codifiers using different criteria for acceptance of a proposition and data sources that do not always overlap, appear to be converging on roughly similar propositions; there is little inconsistency in the propositions advanced. Even more comforting is the growing discovery, at least by American disaster researchers, that students of disasters in other societies, such as in France, Australia, Japan, Belgium, Sweden, and England, are independently arriving at the same rough kinds of generalizations in their codifications. Researchers studying panic behavior, for example, without knowledge of the ideas collated by students in other societies, arrived at surprisingly similar general conclusions about the characteristics, frequency, and conditions associated with panic behavior in disasters (Quarantelli 1977b, Ro- sengren et al 1975). Furthermore, many of the middle-range propositions that are eventuating from the codifications are proving to be consistent with generalizations well established in sociological and social psychological research generally. For example, there has been the "rediscovery" in the disaster area of the relatively secondary importance of the mass media compared to interpersonal sources, particularly with respect to the warning process (Drabek 1969, Drabek & Stephenson 1971), and a "refinding" that being subject to extreme stress in the instance of disasters leads to as little severe psychological breakdown of victims (Bates et al 1963, Taylor 1977b) as was found long ago for wartime bombings. Therefore, while much can be criticized specifically about any of the codification efforts, in general such attempts can be seen as both a mark of and a major help in the recent advancement of the quality of disaster research.

Social Organizational Rather than Social Psychological Emphasis Sociology, as is well known, has generally undertaken far more social psychological than social organizational research. In recent years, however, a far greater social organizational emphasis has come to prevail in the field. In its history, disaster research has closely mirrored this general picture in the discipline.

The early work, particularly that done at the National Opinion Research Center (NORC) at the University of Chicago (see, e.g. Marks et al 1954), clearly had a social psychological emphasis, as did the disaster research of others trained at the




University of Chicago (e.g. Killian 1956) or elsewhere (Moore et al 1964). Much of the early myth-smashing about panic, personal breakdown, antisocial behavior, etc, reflected a similar emphasis. However, as suggested by Fritz, and as reflected in some of the National Academy of Science (NAS)-supported research with which he was associated, a more social organizational emphasis came into being in the early 1960s. This change was strongly reinforced by Barton's (1970) classic work and the conscious commitment to such an orientation by DRC almost from its formation in 1963. Thus in the last decade, although exceptions can be readily found (e.g. Lucas 1966, 1969; Wolf 1975), most of the disaster research by sociologists has been of an organizational nature.

This emphasis provided a useful corrective to the early inbalance in the field in that it led to a seeking of explanations in the social setting or structure, rather than in the psychological makeup of individuals. Fundamental questions about groups as groups, their composition and behavior, and their interrelationships were addressed and not slighted, as in early research efforts. However, the social organizational emphasis in the last decade or so has not been altogether functional for research in the area. For example, the NORC survey study of the 1952 Arkansas tornadoes addressed a massive range of social psychological problems and issues (Marks et al 1954). It is perhaps indicative of the weakening of the social psychological emphasis that this study of over two decades ago still remains the best designed and executed one of its kind even though the published material left unanalyzed most of the gathered field data. More recent surveys, no matter how well done, have tended to restrict themselves to a considerably narrower range of topics (e.g. Wilkin- son & Ross 1970, Bourque et al 1976, Erickson et al 1976). The social organizational emphasis in disaster studies was needed, took hold, and now has a firm base in the overall disaster research enterprise.

Groups Rather than Individuals as Basic Units A related although somewhat separate problem to the social psychology-social organization emphasis just noted is the question of whether the group or the individual is to be taken as the basic unit of description and analysis. Increasingly in disaster research in recent times, the group or the organization has replaced the person as the prime object of attention. This shift gave rise to some fundamental and significant issues of both a theoretical and methodological nature.

Some researchers took seriously the admonition of Fritz some time ago that, at least at the practical and operational level, the effectiveness and efficiency of disaster response is dependent more on the viability of the emergency organizations involved in the crises than it is on the psychological state or readiness of individual victims (Dynes 1975:8). Accordingly, almost all of the range of organizations typically involved in emergency disaster response have come under intensive scrutiny. Thus there are available specific descriptive, as well as more general, accounts of the response of police departments, for example (Kennedy et al 1969, Kennedy 1970), fire departments (Warheit 1970), the utilities (Warheit 1968, Adams 1969, Brouil- lette 1970), hospitals (Drabek 1968, Stallings 1970, Taylor 1974, Blanshan 1975), welfare and social service agencies (Warheit 1968; Ross 1969, 1970; Adams 1970;




Taylor 1976), military groups (Anderson 1968, 1969c), civil-defense agencies (An- derson 1969b, Dynes & Quarantelli 1975), and the mass media (Waxman 1973a,b; Kueneman & Wright 1976; Hannigan 1976). One reaction to many of these diverse studies has been "that trying to construct theory by reviewing data on any single collection of units that share a common title, e.g. the Salvation Army, but which may differ greatly in many other structural characteristics, e.g. size or degree of formalization, may prove to be as much of a pitfall as using. . ." physical features to conceptualize disasters (Mileti et al 1975:76-77; see also Haas & Drabek 1973:362-74).

There appear to be at least three problems that have come to plague researchers on organizations in disasters. For one, a number of the key groups operative in emergencies do not have the classic structural dimensions of formal organizations, particularly of bureaucracies, as Max Weber postulated and as much of sociological organizational theory has assumed since his formulation. Such organizations as the Red Cross, for example, have very amorphous boundaries, few clear-cut members, indistinct subunit relationships, vague lines of authority, and global, almost symbolic, rather than instrumental goals (Dynes 1968; Stoddard 1969; Adams 1970, 1972). Substantively very little of current complex organizational theory seems applicable to such "indistinct" groups, and methodologically it is not always mani- festly clear what should be studied and how.

Second, many organizations and groups operative in emergencies have the characteristics of formal organizations except for those features that depend on the organization's size, which is often not an insignificant factor. These kinds of organizations, such as the mayor's office, the local civil-defense agency, and most mass-media radio and television stations are, in numerical terms, small groups. In fact, in some cases they consist only of one role incumbent, and seldom have more than a dozen members. These groups therefore lack the division of labor, the complex communication patterns, the stratification system, etc, which are suppos- edly the hallmarks of formal organizations. While these kinds of groups are usually easily found and can be readily located for study purposes, they, like the "indistinct" groups mentioned above, nonetheless do not lend themselves substantively and methodologically to assumptions used by most sociological theories of complex or formal organization.

Finally, disaster situations tend to be peopled by emergent groups, entities that had no existence prior to the crisis; these often have only transitory existence, but their functioning may be crucial to the whole trans- and post-disaster response (Bates et al 1963, Quarantelli 1966, 1970; Parr 1969, 1970; Taylor et al 1970; Forrest 1972, 1977). The emergent groups may range from religious interfaith welfare agencies (Ross 1976), to search-and-rescue teams (Zurcher 1968), to temporary overall community-coordinating groups (Dynes 1974:146). Formal organizational theory in sociology is woefully deficient on the question of emergent groups or intermittent organizations, if there is any recognition at all of the phenomenon- an exception would be Etzioni (1961:288-96). In fact, sole dependence on the theory could easily lead to an ignoring of the phenomenon, an orientation assumed by some early disaster researchers interested in complex organizations. In most cases, the




pervasiveness of emergent groups was such that field researchers had to recognize their existence.

The three kinds of groups we have noted that are likely to be encountered in disaster situations have posed problems for researchers. However, on balance the result has probably been salutary. The failure of standard organizational theory to recognize such groups and the weakness of standard organizational methods for studying such groups has generally forced disaster researchers to be innovative. They have either developed new conceptualizations or, as we show below, have reformulated some ideas from existing orientations, such as collective behavior or open-system theories. Methodologically, the problems have forced more ingenuity in the field in locating informants and respondents from such groups, leading, for instance, to experimenting with different kinds of "snowballing" techniques for obtaining data (Forrest 1972).

Increasing Use of "System" Notions As just indicated, sociologists doing research in the disaster area have had to accept the existence of emergent groups. They have also been increasingly inclined to accept "system" notions. However, this incorporation of the idea of system into theory and research has been generated by factors somewhat different from the empirical reality of what field researchers have found. Along one line, a system perspective has been dictated by the beliefs of operational personnel and planners involved in disaster-related tasks and activities that they are manning systems. Along another line, researchers and theorists have tended to see the system perspective as one avenue by which an interdisciplinary attack on disaster problems could be mounted.

In the course of the development of American institutions, the idea has grown that delivery services cannot be provided piecemeal. It is assumed that the services can be delivered only through an integration of a set of organizations, either horizon- tally or vertically linked. Thus, considerable effort has gone into the creation of what are called, for example, mental health service delivery systems, or medical care service delivery systems. Consequently, when a disaster impacts an area, it becomes meaningful, given the assumption of a system, to ask "what services were delivered by the mental health system after the Xenia tornado (Taylor et al 1976)?" Or, as in a study currently under way at DRC, "what emergency medical services are being delivered by the medical care system at times of disasters?" The planners and operational personnel, of course, assume the existence of a system. The question they want answered is "how efficient and effective is the system in delivering the services?" In contrast, however, some researchers, including some sociologists, (Taylor 1976, Wright 1976) have tended to take the given as a question: "was there a delivery system at all, either before or after the disaster?" Despite some sociological challenge of an operational assumption, the overall consequence is that there is increasing use of system notions in disaster research.

From the perspective of other sociologists interested in the disaster area, the notion of system suggests a way of circumventing disciplinary limitations and allowing interdisciplinary research and theory. Thus, in a recent statement, Mileti




et al (1975:147-48) asserted that "the concept of system stress appears to be a mechanism whereby research findings from studies of disaster might, first, be integrated together, and second, linked to other areas of inquiry in the social sciences. General systems paradigms proposed by several theorists could provide a critical integrative link." The statement goes on to further state that "system independence can be maintained, and the consequences of stress within any one systemic level, e.g., group, organization, or community, can be investigated. In much the same way, system interpenetration can be recognized and researched at several systemic levels, such as the effects of psychological stress on biological functioning, or the effects of organizational stress on the power structure in a community." This perspective is seen as allowing disasters to be viewed as events that are influenced by ongoing system processes and that, in turn, may affect those same processes.

Combinations of Collective-Behavior and Complex-Organization Approaches Besides, or perhaps in addition to, the call for a system perspective on disasters, there has been advocacy of a merger between different sociological stances or theoretical perspectives on particular disaster problems. More specifically, the use- fulness of combining a collective-behavior and a complex-organizational perspective has been suggested. The suggestion is part of a much larger trend in sociology; the attempt to meld the long-held traditional distinction between conventional and collective behavior, or as Marx & Wood (1975:365) phrase it: "This trend has continued and accelerated to a point where important aspects of collective and conventional behavior can be conceptualized within common frameworks. The areas of collective behavior and social organization are no longer so distinct." Disaster-research studies appear to have contributed substantially to this kind of thinking (Weller & Quarantelli 1973).

Until recently, theories of collective behavior and of complex organizations have not been viewed as having much in common. Organizational theories have tended to assume that their focus was entities with definite boundaries, rather clearly identified members, formal role positions, established lines of authority, and specific tasks. In contrast, collective-behavior formulations have assumed behavior in the process of coming into being and generally at the opposite end of a continuum from highly institutionalized or bureaucratic behavior. Part of the reason that until recently few direct efforts have been made to reconcile the two approaches is that sociologists were seldom forced to study situations where both forms of behavior were simultaneously present in the immediate situation. Sociologists looking at disaster situations could not avoid such situations; both kinds of behavior abound in such crises (Parr 1969).

Insofar as disasters are concerned, the most direct treatment of this issue is by Dynes & Quarantelli (1968). Drawing on the field studies of DRC, they observe that very often there is, at least in ideal-type terms, a sequential involvement of institutionalized and new groups in community stress situations. That is, established organizations tend to be the first to respond in disasters. They are followed in their response by the involvement of expanding, and then extending, organizations. Only




when these more traditional organizations are participating is it likely that there will be involvement of emergent organizations or groups. More important, it is the very involvement of the more established and traditional organizations in the disaster response that creates the conditions for the generation and emergence of new groups. Thus any attempt to describe and to analyze the overall phenomenon requires that both perspectives, that of collective behavior and of complex organizations, be brought to bear at the same time. At present, the formulation advanced appears to have contributed more to thinking about collective behavior generally than to disaster research. However, the formulation does link a number of different hypotheses, all of which could be put to an empirical test.

Increasing Emphasis on Pre-Impact Period as Source of Post-Impact Changes Some of the early research in the disaster area at times seems to assume that the trans- and post-disaster behaviors exhibited were the almost exclusive results of factors at play in the emergency period. Although probably no researcher would have seriously defended such a position, later writings have openly challenged this as a viable possibility (Brouillette & Quarantelli 1971, Mileti et al 1975). More importantly, disaster research, at both the organizational and the individual level, has increasingly argued for the "principle of continuity," that is, pre-disaster behavior is probably the best indicator of trans- and post-disaster behavior (Quarantelli 1977).

We can note the principle of continuity being advanced in a handful of studies of organizational change that have been conducted in the disaster area (Anderson 1966, 1969, 1972; Blanshan 1975, 1977; Taylor et al 1976; Ross 1977). Not only do the studies generally indicate that relatively little organizational change occurs as a direct result of a disaster, but such changes and shifts as do occur in structure and functions were, in most cases, already manifest in the pre-disaster period. In other words, disasters do not seem to initiate major organizational changes. At most, they appear to accelerate existing trends and, in this sense, reflect the principle of continuity.

Generally speaking, the principle of continuity also applies to the behavior of individuals. Studies of mental health problems, for example, do not show any large increases in such problems after a disaster, especially if problems of mental illness are being examined (Taylor et al 1976, Taylor 1977b). Similarly, as a study of the Wilkes-Barre flood (Wolensky 1975) indicates, if local officials are ineffective prior to a disaster, they are likely to continue to manifest such ineffectiveness after a disaster.

While the principle of continuity as a general principle seems validated by the disaster research so far undertaken, some qualifications should be noted. Drastic personal and social changes may occur under special and often unique conditions (Weller 1973). For example, it appears that the destruction of the community context in the Buffalo Creek flood dam disaster resulted in widespread and severe psychopathology (Erikson 1977). The community context is seldom that badly disrupted in most American disasters, and, consequently, there are relatively few




mental illness problems generated by most disasters (Taylor 1977b). Similarly, some studies (Anderson 1966, Weller 1973, Ross 1977) have found certain organizations changing in various ways after some disasters. But if the likely probability must be summed up in one sentence, it can be said that post-disaster change is unlikely unless pre-disaster trends for change were already operative. In this respect, disasters are different from civil disturbances, at least from the urban riots of the late 1960s, which generated substantial organizational changes, especially in American police departments (Kreps 1971, 1973).

Developing Focus on Functional and Dysfunctional Long-Run Consequences Longer-term consequences of disasters were an object of attention in what was the first genuinely sociological study of disasters, Prince's (1920) study of the Halifax explosion. In fact, the title of the work is Catastrophe and Social Change. However, subsequent studies generally did not return to research on long-run problems until the NAS sponsored work on Hurricane Audrey (Bates et al 1963). This longitudinal study reexamined a community four years after impact, focusing on such matters as whether there was a therapeutic community in the short and in the long run after impact, the mental health effects of the hurricane, and social changes in organizations and the community in the intervening years after impact. It was reported that a therapeutic community did not seem to emerge, that there were few noticeable serious mental health effects, and that relatively little social change occurred.

Since that work there has been an acceleration in the number of studies that have attempted to take a broader time perspective. For example, a very recent study (Western & Milne 1976) in Australia examined, ten months after the disaster, social and psychological consequences to victims of Cyclone Tracy, which hit Darwin. The research essentially found that the cyclone victims who remained in Darwin experi- enced the least disruptive social and psychological effects. On the other hand, those evacuees who still had not returned to Darwin at the time of the study suffered the most effects, with returned evacuees falling in between the other two categories.

Other longitudinal studies are breaking new ground. For example, a recently initiated study at the University of Massachusetts under Peter Rossi (personal communication) is examining how long it takes for communities to recover from disasters of different types, of varying severity, and within communities having different characteristics.

Of interest in the newer studies is the conscious attempt to ascertain both functional and dysfunctional results of disasters. It is no longer being assumed that consequences are necessarily negative. At the individual and household level, both Drabek et al (1973), in their study of the Topeka tornado, and the DRC (Taylor 1977b), in its research on the Xenia tornado, have found that in some cases disasters have been beneficial to the victims involved. For example, Drabek has established that generally victim families are better off with respect to family solidarity and relationships than matched nonvictim families in the same community with which they were compared. In some ways, the disaster strengthened impacted families. Certain ongoing studies at the DRC also show that, both at the organizational and




community levels, there are agencies and sectors that in the long run benefit from disasters. These organizations and segments of the community are better off from an economic, status, or power position than they were before the disaster; in fact, their better positions can be directly traced to activities associated with the disaster. Although in some respects, such findings make a closed circle back to what was long ago suggested by Prince (1920), the later studies rest on a much firmer empirical basis.

At one level it might appear that some of these findings are not consistent with the principle of continuity mentioned earlier. At another level, however, it is obvious that what is being found is that disasters contribute both to stability and to change, just as they generate both consensus and conflict (Quarantelli & Dynes 1976), and that some of the more specific conditions associated with one or the other outcome are now being slowly uncovered by the more current disaster studies. In general, there is continuity from the pre- to the post-disaster state, but under particular conditions certain major changes can occur, and these may be functional or dysfunctional, depending upon the evaluative stance taken. This is certainly a step away from the assumption that disasters were unreservedly bad, which prevailed for years in the disaster area. Sociologists in other areas of study who long ago discovered that "good" can come out of "bad" and vice versa, should have foreseen this also in the disaster area. But whether foreseen or not, perhaps the more important point is that current research is tending to look for functional as well as dysfunctional aspects of disasters for both individuals and groups. This is certainly a change from most of the research of two decades ago that had forgotten the leads provided by Prince, although isolated pieces of research had continued to point in the right direction (e.g. Deutscher & New 1961).

Model Building Model building in the disaster area is not new. Almost four decades ago in one of the very first articles on disasters in the sociological literature, Carr (1932) postulated a time-sequence pattern of differing stages linked to a four-fold typology of disaster events. Various other early attempts to develop models were recently discussed by Stoddard (1968). While most early formulations are rather modest in intent, two attempt to capture a rather full range of disaster phenomena. Schatzman (1960) developed a sequence pattern model of disaster and its consequences for community that emphasized the continuity between disaster and nondisaster behavior. Loomis (1960) fitted the then-existing knowledge about disasters into his Pro- cessually Articulated Structural Model, which was supposed to encompass all sociological phenomena.

More recent model building has tended to take one of two forms. Some disaster students develop very detailed and very complex models with respect to a specific question or problem. A rather sophisticated example of this approach is Mileti's (1974) causal model of warning responses for individuals. It rigorously links, in a tight but specific way, three exogenous variables (income, severity of last experience, and number of children), three intermediate endogenous variables (degree of person- alness of warning communications mode, a degree of warning confirmation behavior, and degree of specificity of warning sought), and a final outcome variable




reflecting degree of adaptive response (from doing nothing to evacuation). Using path analysis, Mileti tested the model with empirical data from two different warning events in Rapid City, one of which eventuated in a flood. Based on a four- warning sequence for each event, the model was estimated eight times, with a number of statistically significant relationships being uncovered under controlled conditions (e.g. the greater the severity of the last flood experience, the lower the confirmation behavior and the lower the adaptive response).

The other kind of model that is developed tends to be more general and applicable to all disasters. Examples of this are the several sensitizing theoretical frameworks used by the DRC, one of which combines elements of open systems and collective- behavior theories. As applied in a study (Taylor 1976) of the delivery of mental health services in the Xenia tornado, the model assumes, in general form, the linkages shown in Figure 1 (the substantive content of each dimension varies depending on the particular event being studied). In general, it was found that the shift from an established to an emergent mental health delivery system in Xenia after the tornado was related to two kinds of post-impact sociocultural conditions that affected that demand-capability ratio within which the established system func- tioned. The conditions had to do with intersystemic factors and the macroenviron- mental setting. The specific characteristics manifested by the emergent mental health delivery system were the result of the feedback from the interplay between

T The Socioevolutionary Contexts

M E Demand-Capability Ratio

N Components, Domain, and Boundaries of the

E Established Mental Health Delivery System

E~~~~~ Disaster Sociocultural Conditions Impact

T

M Demand-Capability Ratio

E

T Components, Domain, and Boundaries of the W Emergent Mental Health Delivery System 0

Figure 1 Typical DRC theoretical frameworks.




the internal dynamics of the established system and external factors impinging on it after the disaster. The existing delivery system could not respond to the negative feedback it was receiving and eventually it ceased to attempt to adapt. At the same time, the positive feedback to the emergent mental health system led it to continue to develop in the direction it was moving.

The substantive trends discussed above naturally do not represent the only changes in the last 10-15 years within the sociologically oriented sector of theory and research in the disaster area. Neither, as implied and at one point noted, are the trends or their results totally consistent with one another. They do, however, represent major changes in substantive orientation that are going on in the area and that, along with some structural trends we discuss next, are setting the tone for those sociologists who will be involved in future disaster research and theory.

STRUCTURAL TRENDS Disaster Research and Public Policy and Agency Responsibility Until recently, research funding for disaster studies in the United States has been somewhat erratic and discontinuous. In fact, for years the major impetus for social science interest in disasters at the federal level was provided by a handful of sociologists involved with the NAS. These sociologists first headed the Disaster Research Group in the late 1950s, served on the Advisory Committee on Emergency Planning and related Academy committees, and are currently involved in the descendant of these earlier groups, the Contingency Planning and Emergency Services Group. Playing a particularly key role in all these groups has been Charles Fritz, a sociologist whose status as a researcher plus knowledge of the scientific community and federal agencies has allowed him to play an important mediating role in linking the research community to potential users and possible sponsors of research.

Most federal agencies have actually been research consumers rather than research supporters. Furthermore, since disaster responsibility is diffuse on the federal level, such interest as there has been in supporting disaster research has also been diffuse. This interagency diffusion is additionally complicated by the understandable tendency of federal technical personnel to favor research in disciplines similar to their own training, e.g. engineers like to support engineering research. For a long time, therefore, sociologists, not well represented in the higher reaches of most government agencies, had difficulty "selling" their approach and skills to such agencies and departments, which support disaster research. Those agencies favorable to sociological research, such as the old Office of Civil Defense and the National Institute of Mental Health, were very few in number.

However, there have been more recent trends at the federal level that have generated even greater interest in and some additional support for disaster research. Some of these trends, such as the increased governmental support of research and development and the greater role assigned to the National Science Foundation (NSF), have had little to do directly with the content of research. Nevertheless, the increased role of NSF has allowed it to develop its support of applied research, and a newly developed program in the disaster area has been put under the direction of




a sociologist with extensive field experience in studies of disasters. Other trends at the federal level have had an even more direct effect on the disaster area. Govern- mental concern with handling "emergencies" of all kinds was reinforced by the civil disturbances of the late 1960s and led to some turning to sociological research for answers. The increasing role of the federal government, often mandated by legisla- tion, in the delivery of disaster services at the local level has made some agencies quite interested in what sociologists could offer with respect to both policy and effectiveness in such delivery.

Even when disaster research was not directly supported, some trends have helped to turn governmental agencies to sociology for advice and recommendations. Thus, a probable technological breakthrough in earthquake-prediction technology led to the appointment of an NAS Panel on the Public Policy Implications of Earthquake Prediction, which was chaired by a former president of the American Sociological Association and included two other sociologists as panel members (Turner 1976). Uncertainty, both as to the policies that should be followed and the effectiveness of the aid provided, has led the Agency for International Development to ask the NAS to appoint a committee to examine the whole matter of United States assistance in foreign disasters. This committee on International Disaster Assistance is also headed by a sociologist.

Thus a combination of factors has led to policy concerns and questions of effectiveness of actions undertaken by agencies involved in the federal response to disasters. A result has been an interest in policy and general research in the disaster area. Sociologists have increasingly become involved in such research.

Institutionalization of Disaster Studies in Academic Settings Until the early 1960s, disaster research, while generally undertaken at scattered academic institutions, was rather discontinuous since it was comprised primarily of isolated research projects. The first ongoing academic research operation came into being with the establishment of the DRC at the Ohio State University in 1963. Since the Center, from its inception, was funded from several sources and for extended periods of time, it was able to maintain field teams ready to do very rapid nationwide (and for a time worldwide) research in crises and to accumulate a data base as well as specialized library resources. Tied in to a major sociology teaching department, DRC was also able to provide experience in disaster research for dozens of graduate students. One consequence is that the vast majority of currently active sociological researchers in the United States and Canada who have a major interest in social and behavioral aspects of disasters have been at some time DRC staff members or have been trained by sociologists who themselves came out of the DRC.

Another concentrated program for disaster research was initiated at the Univer- sity of Colorado in 1966 by a former DRC Co-Director. Within the Institute of Behavioral Science, work on warning systems and social consequences of weather modification led eventually to a massive study on the assessment of research on natural hazards (White & Haas 1975). The Colorado work has been continuously interdisciplinary, with sociologists playing a prominent role, although little has been attempted by way of systematically teaching graduate students. The Institute has




produced a series of publications of an interdisciplinary and applied nature, (e.g. Mileti 1975) and its personnel have undertaken a number of field trips to disasters outside of the United States.

While DRC and the Colorado group remain the only continuous ongoing disaster-research clusters in America, sociologists have become increasingly involved in disaster studies at other universities. Thus, research on social and behavioral aspects of disasters has been or is being undertaken by prominent sociologists at UCLA, and the Universities of Kansas, Delaware, Massachusetts, Minnesota, Pittsburgh, Arizona State, and Iowa State, among other universities. While the Institute of Disaster Preparedness at the University of Southern California has been established in the public administration area, its research director is a sociologist. Additionally, disaster research is proceeding at three major Canadian universities under the direction of sociologists trained at the DRC.

Finally, sociological research into disasters has been initiated in the last 10-15 years in other countries around the world, and sociologists are prominent in these studies in Belgium, England, Australia, Japan, France, and New Zealand. One recently established research center in Italy is an integral part of a sociology department, as is another group in Sweden.

At the research level, the last decade has seen disaster studies becoming firmly embedded in academic settings around the world. For a variety of reasons, the very fact that sociologists are prominent in this development has raised both the quality and quantity of the studies being undertaken. Another important consequence of disaster research within an academic framework is the steady stream of new graduate students systematically exposed to the area.

Emergence of a "Critical Mass" of Social and Behavioral Scientists Interested in Disaster Research In the early days of disaster research in the United States, an isolated professor or two labored with a graduate student or two. Those involved in the research, furthermore, had little outlets for their work; indicative of this is the fact that the classic NORC work on the Arkansas tornado (Marks et al 1954) never received regular publication, but was released only as a final report for the sponsoring agency. There was little chance of the disaster researcher quickly becoming aware of other works in the area. There simply was not the mass of researchers interested in the topic, a critical factor for the development of the field.

The situation has changed drastically. Several contrasts can be made to indicate the growth of this critical mass of researchers interested in the disaster area. Prior to 1960, there were no more than 20 sociologists who had had any direct field experience in disaster research. By the mid- 1970s, this number had increased 8-10- fold, much more rapidly than the expansion of the profession itself. Another mea- sure is the production of dissertations on disaster-related topics. Starting with Prince's (1920) classic study, 13 such dissertations were written by 1960, but from 1961 to 1976 there were 34 dissertations produced, almost all in sociology. Sections devoted to discussion of disaster research in progress have been organized at the




American Sociological Association national meetings, at several regional meetings, and at the last World Congress of Sociology.

The number of publications has increased dramatically in standard sociological journals, in interdisciplinary journals, and through in-house reports from the various research centers around the world. A whole issue of the American Behavioral Scientist (Quarantelli & Dynes 1970a) was devoted exclusively to disaster-related articles. The International Sociological Association is sponsoring a forthcoming volume Disasters: Theory and Research edited by Quarantelli (1977a), which will contain a set of previously unpublished papers at the forefront of theory and research in the disaster area. Two new professional journals have appeared that will report on disaster research; one, Mass Emergencies: An International Journal of Theory, Planning and Practice, while publishing articles from all disciplines, is under the editorship of two sociologists, Nehnevajsa and Quarantelli. There are several newsletters in the disaster area published by research centers in which sociologists are involved: Unscheduled Events, issued by DRC since 1967; Disaster Research, put out by the University of Bradford Disaster Research Unit in England; and the Natural Hazards Observer from the Institute of Behavioral Science at the University of Colorado.

The number of sociologists involved in disaster research and the existing information-exchange sources means that an international network of disaster researchers is now in existence. This manifests itself in specialized meetings; for example, in 1972 there was a week-long joint Japan-United States Research Seminar held at DRC involving several dozen researchers, most of them sociologists (see Proceedings 1972). More recently, another international meeting was held in Paris at the Center for Psychosociological Study of Catastrophes (Comportements 1975). Further international meetings of disaster researchers are being planned soon in Japan and Italy, and the possibility ofjoint cross-cultural studies is being explored around the world.

GAPS AND CHALLENGES-DIRECTIONS FOR THE FUTURE

Given the trends just discussed, where does disaster research stand today? In what directions should it go? The present-day standing of the field has been discussed in previous pages. Here it is sufficient for us to make only a few more general points before indicating some possible directions for the future.

While much of what has been undertaken in disaster studies has been of a descriptive and inventory nature and devoid of direct sociological relevance, much of the sociological research that has been conducted has been guided implicitly, if not explicitly, by middle-range theories. For example, studies of emergent groups have used collective behavior and open-system theories. Many warning studies have assumed a symbolic interactionist framework, although this has not always been recognized by the researchers involved. Research into complex organizations has not infrequently taken one of the several available organization theories.

In general terms, the quality of research would probably improve if the implicit theoretical frameworks used were made more explicit. Perhaps more important, new




theoretical orientations should be brought to bear. An excellent example of the fruitfulness of what can be done is illustrated by Turner (1967), taking the recurring observation about the social solidarity that typically follows disasters, reinterpreting it in terms of Durkheim's principle about mechanical and organic solidarity, and then showing parallel applications of the same principle to quite disparate phenomena, from romantic love to nationalism.

In the main, there have been no startling theoretical breakthroughs in the empirical studies, even in the accelerated research in the last decade or so, indicating that not enough of the necessary spade work has been done. We suggest that for the future there is a need to (a) empirically establish new generalizations, (b) empirically confirm what is merely suspected, and (c) empirically question the doubtful and the dubious. An effort then should be made to avoid luring pitfalls and enticing traps, within or outside sociology, no matter how fashionable they may be.

In one sense, this is a partial call to do more of what is being done in disaster research. This would, of course, not be necessary if we had solid sociological knowledge of the total range of disaster phenomena. There are whole areas and questions about which empirical data is totally or almost completely lacking. The general failure of all codification efforts, so far, to surface significant theoretical questions that were overlooked because the several empirical findings were not reexamined within a single context, also strongly suggests that we need to do considerable groundwork. We may be close to significant theoretical advances along some lines (see Kreps 1977), but we still need to accumulate solid empirical accounts in a variety of areas.

Whole institutional sectors have been ignored by sociologists in disaster research. About the only empirical work of any sociological value in the political arena is a modest paper by two political scientists (Abney & Hall 1966) and some first-class research by an Australian political scientist (Wettenhall & Power 1969, Wettenhall 1975). Brown & Goldin (1973) may argue for the inherently political nature of all disaster phenomena, but no one seems to have heard them. The business world has also been ignored; there is not even a single descriptive account of the business corporation, or of labor unions, in disasters. Apart from some unpublished DRC material on banks and savings and loan institutions, the whole financial complex of American society with respect to disasters has not been touched sociologically despite its obvious importance in such matters as community recovery and rebuild- ing. Contrary to the anecdotes about the relationship of the New York city blackout and a rise in the birthrate, we know next to nothing about the consequences of disasters on sexual-erotic behavior.

Even in areas where substantial knowledge has been accumulated, there are surprising gaps. Solid data has been gathered about the response of individuals to warnings, but the processing of warnings through emergency organizations has barely been touched. Much is being learned about hospitals and the delivery of emergency medical services to disaster victims, but the handling of the dead remains an almost unknown topic. There are only two known empirical studies on the question (Hershiser & Quarantelli 1976, Pine 1974). Still other voids of knowledge could be indicated and some have been suggested elsewhere (Taylor & Quarantelli




1977), but we believe that enough illustrations have been given to emphasize the range of sociologically relevant research lacunae that ought to be filled.

Apart from establishing new generalizations, we need to confirm empirically what is suspected on the basis of already-conducted studies that fall short of being definitive, much less conclusive. As we noted earlier, many "findings" that have become embedded in the research tradition are based on one or two studies conducted at specific time periods in specific stiuations. Since researchers usually like to attempt something new, replication is often not attempted in the disaster area. In addition, the occurrence of certain phenomena in many different situations is often accepted as given, rather than being seen as worthy of future investigation.

Thus, there has been some work done, much of it in the early days of disaster research, on the convergence phenomena (Fritz & Mathewson 1957), blame assign- ment in disasters (Bucher 1957, Drabek & Quarantelli 1967), the reaction of children to major mass emergencies (Perry & Perry 1959), and disaster subcultures (Moore et al 1964, Anderson 1965). However, later confirming work has not been done. The findings of the early research have been generally accepted. Many other illustrations could be given, but our point is that additional empirical research is needed to replicate and strengthen what is already known, and to empirically pin down what is strongly suspected but not yet definitely proved.

Much of the sociological disaster research undertaken so far has effectively under- mined certain widespread public beliefs about behavior in disasters. In fact, it has even been empirically established through surveys that the public holds mistaken notions (Wenger et al 1975, S. Blanshan, personal communication). However, there is some danger that in attacking widespread public misconceptions, overstatements may be made. The evidence, for example, seems fairly solid that looting is a very rare phenomenon in American disasters (Wright 1977), but data supporting the absence of looting in all societies is lacking. In fact, there are some impressionistic observations which suggest that looting may occur in disasters in other societies where pre-disaster norms about the security of property differ from those in the United States (Kates et al 1973). It may still turn out that looting is very rare in all disasters, but at the moment the evidence primarily supports the absence of looting in American society and perhaps in some other Western countries.

Another example can be cited from the mental health area. There is fairly solid evidence that there is almost no likelihood of mental illness in the wake of the typical American disaster. However, there is some scattered work suggesting that the truly massive catastrophe may have somewhat different consequences on mental health than the typical, moderately sized disaster in American society. At least Erikson's (1977) brilliant sociological analysis of the destruction of the community fabric in the Buffalo Creek dam disaster suggests that a qualification might be added to the now widely accepted generalization that mental illness does not eventuate from disasters. Neither the general proposition nor the qualifier can be established unless replicatory work is done, especially in somewhat different kinds of social contexts.

One of the weaknesses of disaster research has been its overwhelming reliance on studies conducted primarily in the United States by American sociologists (Taylor 1977c). It may turn out that studies conducted by non-Americans in other than




American disaster settings may produce the same results. This has already occurred in, for example, an English study of a plant explosion (Westgate 1975), a Japanese study of panic in a nightclub fire (Abe 1976), and a Philippine study of family behavior in evacuation (Carroll & Parco 1966), which did not challenge, in any way, the major observations made in similar American disasters. We ourselves believe that there are many principles that are universal and cut across different cultural settings (Quarantelli 1977a). But a belief is not evidence, and too much in the disaster research literature rests on belief rather than evidence. More general cross- cultural studies, as advocated by us (Taylor & Quarantelli 1977) and others, would give greater confidence in what we so often assert about what sociological research has established about human and social behavior in disasters.

In this paper we have somewhat slighted nonsociological disaster research and some of the research by sociologists in interdisciplinary contexts. This is not the result of any illusion that the sociological perspective is the only way, or even the only valid way, to look at disasters. However, our selective treatment and analysis is based on the strong conviction that sociology does have specific and distinctive contributions to make to disaster studies. We stress the value of this orientation because we believe that two currently fashionable notions threaten to undermine the substantial and important potential of what sociology can offer to theory and research in the disaster area.

There are recurrent cycles of enthusiasm for interdisciplinary research in the disaster area. A peak of such enthusiasm engulfed the early period of disaster research in the middle 1950s. Federal agencies in the United States government that provide research funding are leading exponents of the current wave of enthusiasm for interdisciplinary research to which the social sciences generally are now being subjected. As sociologists ought to at least recognize, the insistence on interdisciplinary efforts is a means of resolving competing claims of disciplines for scarce research funds, and is also caused in part by the understandable inability of funding agencies to reconcile seemingly different answers provided by different disciplines to what appear to be the same practical and operational questions.

Of course, there is a place for interdisciplinary research. In fact, such a setting is a good testing ground for what different disciplines can provide by way of concrete answers to often difficult, but usually very important, practical questions. However, the past history of interdisciplinary research, including efforts in the disaster area, is not supportive of the ideas that better research results are obtained or that applications of findings are more easily accomplished by taking an interdisciplinary stance. In such an approach, contributions of different disciplines are often reduced to the lowest common denominator, which is sometimes only slightly, if at all, above a common-sense level. Research findings in an interdisciplinary setting are often even more difficult to translate into practical terms. While obviously some sociologists should take the interdisciplinary path in disaster research, others, for the good of the discipline as well as for the good of those whom the discipline serves, should continue to do high-quality sociological research on important questions regarding disasters.

Additionally complicating any attempt to follow this admonition is the current seductive lure of "policy-oriented research." Policy-oriented research usually means




that administrative agencies have the right to determine the research priorities of disciplines, both as to substantive questions and methodological procedures. In an area such as disaster research, where findings could have direct policy implications, it becomes very difficult for disciplinary researchers to maintain some degree of autonomy. Furthermore, to accept policy implications as the guide to research, of course, provides no assurance that research findings will have impact on future policy. Such research has not had impact in the past, and there is no reason to suggest that it will in the future. Perhaps the best sociological policy is for sociologists to do as they generally have in much past disaster research, that is, to set their own disaster-research priorities. This position has paid off in the past, as we have tried to demonstrate in the previous pages on what sociological research has produced by way of knowledge of social response to crisis and disaster.

ACKNOWLEDGMENTS

We want to thank past and present members of the DRC staff who, by their questions and inquiries, helped us to recognize many of the problems and issues we discuss in the article. We want to acknowledge, in particular, the very conscientious editing by Miriam Morris and the very able typing by Ruth Chalfant of the manu- script of this paper.

Literature Cited

Abe, K. 1976. The behaviour of survivors and victims in a Japanese nightclub fire. Mass Emergencies 1: 119-24

Abney, F. G., Hill, L. B. 1966. Natural disasters as a political variable: The effect of a hurricane on an urban election. Am. Polit. Sci Rev. 60:974-81

Adams, D. S. 1969. Emergency Actions and Disaster Reactions: An Analysis of the Anchorage Public Works Department in the 1964 Alaskan Earthquake. Colum- bus: Ohio State Univ. Disaster Res. Cent. Monogr. Ser. No. 5. 145 pp.

Adams, D. S. 1970. The American Red Cross: Organizational sources of operational problems. Am. Behav. Sci. 13: 392403

Adams, D. S. 1972. Goal and structural suc- cession in a voluntary association: A con- structed type of the American Red Cross chapter in natural diasters. PhD thesis. Ohio State Univ., Columbus. 190 pp.

Anderson, W. A. 1965 Some Observations on a Disaster Subculture: The Organiza- tional Response of Cincinnati, Ohio, to the 1964 Flood. Columbus: Ohio State Univ. Disaster Res. Cent. Res. Note No. 6. 25 pp.

Anderson, W. A. 1966. Disaster and organizational change: A study ofsome of the long-term consequences of the March 27,

1964, Alaska earthquake. PhD thesis. Ohio State Univ., Columbus. 250 pp.

Anderson, W. A. 1968. Military-Civilian Re- lations During Disaster Operations. Co- lumbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 5. 69 pp.

Anderson, W. A. 1969a. Disaster warning and communication processes in two communities. J. Commun. 19:92-104

Anderson, W. A. 1969b. Local Civil Defense in Natural Disaster: From Office to Or- ganization. Columbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 7. 62 pp.

Anderson, W. A. 1969c. Social structure and the role of the military in natural disasters. Soc. Sci. Res. 53:242-53

Anderson, W. A. 1972. Disaster Research Center studies of organizational change. Proc. Jpn. -US Disaster Res. Semin., Co- lumbus, Ohio, 1972, pp. 74-89

Barkun, M. 1974. Disaster and the Mil- lenium, New Haven: Yale Univ. Press. 243 pp.

Barton, A. H. 1970. Communities in Disaster, Garden City, NY: Anchor, Doubleday. 352 pp.

Bates, F. L., Fosleman, C. W., Parenten, V. J., Tracy, G. S. 1963. The Social and Psychological Consequences of a Natural Disaster-A Longitudinal Study of Hur-




ricane Audrey. Washington DC: Natl. Res. Counc. Natl. Acad. Sci. Disaster Study No. 18. 190 pp.

Blanshan, S. 1975. Hospitals in "Rough Wa- ters": The effects of a flood disaster on organizational change. PhD thesis. Ohio State Univ. Columbus 187 pp.

Blanshan S. 1977. See Quarantelli 1977a Bourque, L. B., Cherlin, A., Reeder, L. G.

1976. Agencies and the Los Angeles earthquake. Mass Emergencies 1: 217-28

Brouillette, J. R. 1970. The department of public works: Adaptation to disaster de- mands. Am. Behav. Sc! 13:369-79

Brouillette, J. R., Quarantelli, E. L. 1971. Types of patterned variation in bureaucratic adaptations to organizational stress. Sociol. Inq. 41:36-46

Brown, M., Goldin, A. 1973. Collective Be- havior: A Review and Reinterpretation of the Literature, Pacific Palisades, Calif: Goodyear. 349 pp.

Bucher, R. 1957. Blame and hostility in disaster. Am. J. Sociol. 62:467-75

Carr, L. J. 1932. Disaster and the sequence- pattern concept of social change. Am. J. Sociol. 38:207-18

Carroll, J. J., Parco, S. 1966. Social Organiza- tion in a Crisis Situation: The Taal Disaster. Manila: Philipp. Sociol. Soc. 59 pp.

1975. Les Comportements Associe's Aux Ca- tastrophes, Paris, 1974, 338 pp.

Deutscher, I., New, P. 1961. A functional analysis of collective behavior in a disaster. Sociol. Q. 7:21-36

Drabek, T. E. 1968. Disaster in Aisle 13: A Case Study of the Coliseum Explosion at the Indiana State Fairgrounds, October 31, 1963. Columbus: Ohio State Univ. Disaster Res. Cent. Monogr. Ser. No. 1. 187 pp.

Drabek, T. E. 1969. Social processes in disaster: Family evacuation. Soc. Probl. 16: 336-49

Drabek, T. E., Key, W. H., Erickson, P. E., Crowe, J. L. 1973. Longitudinal Impact ofDisaster on Family Functioning. Den- ver: Univ. Denver Dept. Sociol. 43 pp.

Drabek, T. E., Erickson, P. E., Crowe, J. L., Key, W. H. 1975. Impact of disaster on kin relationships. J. Marriage Family 37:481-94

Drabek, T. E., Key, W. H. 1976. The impact of disaster on primary group linkages. Mass Emergencies 1:89-106

Drabek, T. E., Quarantelli, E. L. 1967. Scape- goats, villians and disasters. Trans- action 67:12-17

Drabek, T. E., Stephenson, J. J. 1971. When disaster strikes. J Appl. Soc. Psychol. 1: 187-203

Dynes, R. R. 1968. The Functioning of Ex- panding Organizations in Community Disasters. Columbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 2. 86 pp.

Dynes, R. R. 1974. Organized Behavior in Disaster. Columbus: Ohio State Univ. Disaster Res. Cent. Monogr. Ser. No. 3. 235 pp.

Dynes, R. R. 1975. The comparative study of disaster: A social organizational approach. Mass Emergencies 1:21-31

Dynes, R. R., Quarantelli, E. L. 1968. Group behavior under stress: A required convergence of organizational and collective behavior perspectives. Sociol. Soc. Re& 52:416-29

Dynes, R. R., Quarantelli, E. L. 1975. The Role of Local Civil Defense in Disaster Planning. Columbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 16. 105 pp.

Dynes, R. R., Quarantelli, E. L. 1976. Orga- nizational Communications and Deci- sion Making in Crises Columbus: Ohio State Univ. Disaster Res. Cent. Misc. Rep. No. 18. 58 pp.

Erickson, P. E., Drabek, T. E., Key, W. H., Crowe, J. 1976. Families in disaster: Patterns of recovery. Mass Emergencies 1:203-16

Erikson, K. T. 1976. Everything in its Path, New York: Simon & Schuster

Etzioni, A. 1961. A Comparative Analysis of Complex Organizations. Glencoe, III: Free Press. 366 pp.

Forrest, T. R. 1972. Structural differentiation in emergent groups. PhD thesis. Ohio State Univ., Columbus. 197 pp.

Forrest, T. R. 1977. Group Emergence in Disasters, ed. E. L. Quarantelli. Lon- don: Sage. In press

Fritz, C. E. 1961. In Social Problems, ed. R. Merton, R. Nisbet, pp. 651-94. New York: Harcourt, Brace & World. 754 pp.

Fritz, C. 1968. In International Encyclopedia of Social Sciences, ed. D. Sills, 4:202-7. New York: Macmillan

Fritz, C. E., Mathewson, J. H. 1957. Conver- gence Behavior in Disasters: A Problem in Social Control. Washington DC: Natl. Res. Counc. Natl. Acad. Sci. Disaster Study No. 9. 99 pp.

Haas, J. E., Drabek, T. E. 1973. Complex Organizations: A Sociological Perspec- tive, New York: Macmillan. 416 pp.




Hannigan, J. A. 1976. Municipal reorganiza- tion and crisis management agencies: The impact of regional government in Ontario on emergency and protective services. PhD thesis. Ohio State Univ., Co- lumbus. 249 pp.

Hershiser, M., Quarantelli, E. L. 1976. The handling of the dead in a disaster. Omega 7(3):195-208

Hultaker, 0. E. 1976. Evakuera. Uppsala, Sweden: Uppsala Univ. Disaster Study No. 2. 71 pp.

Janowitz, M. 1968. Social Control of Es- calated Riots. Chicago: Univ. Chicago Cent. Policy Study. 98 pp.

Kates, R. W., Haas, J. E., Amaral, D. J., Olson, R. A., Ramos, R., Olson R. 197j. Human impact of the Managua earthquake. Science 182:981-90

Kennedy, W. C. 1970. Police departments: Organization and tasks in disasters. Am. Behav. Sci. 13:354-61

Kennedy, W. C., Brooks, J. M., Vargo, S. M. 1969. The Police Department in Disas- ter Operations. Columbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 6. 56 pp.

Killian, L. M. 1956. A Study of Response to the Houston, Texas Fireworks Explo- sion. Washington DC: Natl. Res. Counc., Natl. Acad. Sci. Disaster Study No. 2. 25 pp.

Kreps, G. A. 1971. Innovation in crisis relevant organizations: A model of the process of organizational change. PhD thesis. Ohio State Univ., Columbus. 223 pp.

Kreps, G. A. 1973. Change in crisis-relevant organizations: Police departments and civil disturbances. Am. Behav. Sci. 16:356-67

Kreps, G. A. 1977. In Disasters: Theory and Research, ed. E. L. Quarantelli. Lon- don: Sage. In press

Kueneman, R. M., Wright, J. E. 1976. News policies of broadcast stations for civil disturbances and disasters. Journalism Q. 53:670-77

Loomis, C. P. 1960. Social Systems, New York: Van Nostrand. 349 pp.

Lucas, R. A. 1966. The influence of kinship upon perception of an ambiguous stimulus. Am. Sociol. Rev. 31:227-36

Lucas, R. A. 1969. Men in Crisis: A Study of a Mine Disaster, New York: Basic Books. 335 pp.

Marks, E. S., Fritz, C. E. 1954. Human Reac- tions in Disaster Situations. Chicago: Univ. Chicago, Natl. Opin. Res. Cent. 861 pp.

Marx, G., Wood, J. 1975. Strands of theory and research in collective behavior. Ann. Rev. Sociol. 1:363-428

McLuckie, B. 1970. The Warning System in Disaster Situations: A Selective Analysis. Columbus: Ohio State Univ. Disaster Res. Cent. Rep. Ser. No. 9. 69 pp.

Mileti, D. S. 1974. A normative causal model analysis of disaster warning response. PhD thesis. Univ. Colo., Boulder. 187 pp.

Mileti, D. S. 1975. Natural Hazard Warning Systems in the United States: A Research Assessment. Boulder, Colo: Univ. Colo. Inst. Behav. Sci. 97 pp.

Mileti, D. S., Drabek, T. E., Haas, J. E. 1975. Human Systems in Extreme Environ- ments: A Sociological Perspective. Boul- der, Colo: Univ. Colo., Inst. Behav. Sci. 165 pp.

Moore, H. E., Bates, F., Alston, J., Fuller, M., Layman, M., Mischer, D., White, M. 1964.... And the Winds Blew. Aus- tin, Tex: Univ. Tex., Hoss Found. Ment. Health. 221 pp.

Parr, A. R. 1969. Group emergence under stress: A study of collective behavior during the emergency period of community crisis. PhD thesis. Ohio State Univ., Columbus. 275 pp.

Parr, A. R. 1970. Organizational response to community crises and group emergency. Am. Behav. Sci. 13:423-29

Perry, H., Perry, S. 1959. The Schoolhouse Disaster: Family and Community as Determinants of the Child's Responses to Disaster. Washington DC: Natl. Res. Counc. Natl. Acad. Sci. Disaster Study No. 11. 66 pp.

Pine, V. R. 1974. Responding to Disaster, Milwaukee, Wis: Buffin. 190 pp.

Prince, S. 1920. Castastrophe and Social Change, London: Kind & Son. 151 pp.

1972. Proc. Jpn.-USSemin. Organ. Commu- nity Responses Disasters. Columbus: Ohio State Univ., Disaster Res. Cent. Monogr. Ser. No. 8. 313 pp.

Quarantelli, E. L. 1966. In Symposium in Emergency Operations, ed. R. Brictson, pp. 3-19. Santa Monica, Calif: Syst. Dev. Corp. 113 pp.

Quarantelli, E. L. 1970. In Human Nature and Collective Behavior, ed. T. Shi- butani, pp. 111-23. Englewood Cliffs, NJ: Prentice-Hall. 232 pp.

Quarantelli, E. L., ed. 1977a. Disasters: The- ory and Research, London: Sage. In press

Quarantelli, E. L. 1977b. In Human Response to Tall Buildings, ed. D. J. Conway.




Stroudsburg, Penn: Dowden, Hutchin- son & Ross. In press

Quarantelli, E. L., Dynes, R. R. 1970a. Edi- tor's introduction. Am. Behav. Sci 13:325-30

Quarantelli, E. L., Dynes, R. R. 1970b. Proc. Nati. Symp. Law Enforcement Sci Technol., Jrd, pp. 323-27. Chicago: ITT Res. Inst.

Quarantelli, E. L., Dynes, R. R. 1970c. Prop- erty norms and looting: Their patterns in community crises. Phylon 31:168-82

Quarantelli, E. L., Dynes, R. R. 1976. Com- munity conflict: Its absence and its pres- ence in natural disasters. Mass Emer- gencies 1:139-52

Rosengren, K., Arvidson, P., Sturesson, D. 1975. The Barseback panic. Acta Sociol. 18:303-21

Ross, G. A. 1976. The emergence and change of organization-sets: An interorganizational analysis of ecumenical disaster recovery organizations. PhD thesis. Ohio State Univ., Columbus. 123 pp.

Ross, G. A. 1977. See Quarantelli 1977a Ross, J. L. 1969. The Salvation Army: Its

Structure, Operation and Problems in Disaster. Columbus: Ohio State Univ., Disaster Res. Cent. Rep. Ser. No. 8. 61 pp.

Ross, J. L. 1970. The Salvation Army: Emer- gency operations. Am. Behav. Sci. 13: 404-14

Schatzman, L. 1960. A sequence pattern of disaster and its consequences for community. PhD thesis. Ind. Univ., Bloo- mington. 164 pp.

Stallings, R. A. 1970. Hospital adaptations to disaster. Hum. Org. 29:294-302

Stallings, R. A. 1971. Communications in Natural Disasters. Columbus: Ohio State Univ., Disaster Res. Cent. Rep. Ser. No. 10. 49 pp.

Stallings, R. A. 1973. The community context of crisis management. Am. Behav. Sci. 16:312-25

Stallings, R. A. 1977. See Quarantelli 1977a Stoddard, E. R. 1968. Conceptual Models of

Human Behavior in Disaster. El Paso, Tex: Tex. West. Press. 148 pp.

Stoddard, E. R. 1969. Some latent consequences of bureaucratic efficiency in disaster relief. Hum. Organ. 28:177-89

Taylor, J. B., Zurcher, L. A., Key, W. H. 1970. Tornado: A Community Responds to Disaster Seattle: Univ. Wash. Press. 118 pp.

Taylor, V. A. 1974. Hospital Emergency Fa- cilities in a Disaster. Columbus: Ohio State Univ. Disaster Res. Cent. Prelim. Pap. No. 11. 13 pp.

Taylor, V. A. 1976. Delivery of mental health services in the Xenia tornado: A collective behavior analysis of an emergent system response. PhD thesis. Ohio State Univ., Columbus. 270 pp.

Taylor, V. A. 1977a. Delivery of Emergency Medical Services in Disasters. Colum- bus: Ohio State Univ. Disaster Res. Cent. Prelim. Pap. No. 31. 25 pp.

Taylor, V. A. 1977b. People in disasters: How do they really react? Psychol. Today. In press

Taylor, V. A. 1977c. See Quarantelli 1977a Taylor, V. A., Quarantelli, E. L. 1977. In

Proc. Nat. Hazards Symp., Canberra, Aust., 1976. In press

Taylor, V. A., Ross, G. A., Quarantelli, E. L. 1976. Delivery ofMentalHealth Services in Disasters: The Xenia Tornado and Some Implications. Columbus: Ohio State Univ. Disaster Res. Cent. Mo- nogr. Ser. No. 11. 328 pp.

Turner, R. H. 1967. Types of solidarity in the reconstituting of groups. Pac. Sociol. Rev. 10:60-68

Turner, R. H. 1976. Earthquake prediction and public policy: Distillations from a National Academy of Sciences report. Mass Emergencies 1:179-202

Warheit, G. J. 1968. The impact of major emergencies on the functional integration of four American communities. PhD thesis. Ohio State Univ., Colum- bus. 270 pp.

Warheit, G. J. 1970. Fire departments: Oper- ations during major community emergencies. Am. Behav. Sci. 13:362-68

Waxman, J. J. 1973a. An analysis ofcommer- cial broadcasting organziations during flood disasters. PhD thesis. Ohio State Univ., Columbus. 225 pp.

Waxman, J. J. 1973b. Local broadcast gate- keeping during natural disaster. Jour- nalism Q. 50:751-58

Weller, J. M. 1973. Organizational Innova- tion in Anticipation of Crisi& Columbus: Ohio State Univ., Disaster Res. Cent. Rep. Ser. No. 14. 109 pp.

Weller, J. M., Quarantelli, E. L. 1973. Ne- glected characteristics of collective behavior. Am. J. Sociol. 79:665-85

Wenger, D. E., Dykes, J. D., Sebok, T. B., Neff, J. L. 1975. It's a matter of myths: An empirical examination of individual insight into disaster response. Mass Emergencies 1:3346

Wenger, D. E., Parr, A. R. 1969. Community Functions Under Disaster Conditions. Columbus: Ohio State Univ., Disaster Res. Cent. Rep. Ser. No. 4. 149 pp.




Western, J., Milne, G. 1977. In Proc. Nat. Hazards Symp., Canberra, Aust.

Westgate, K. 1975. Fixborough: An Analysis of the Human Response. Bradford, Engl: Univ. Bradford, Disaster Res. Unit Occas. Pap. No. 7. 26 pp.

Westgate, K., O'Keefe, P. 1976. Some Defini- tions of Disaster. Bradford, Engl: Univ. Bradford, Disaster Res. Unit Occas. Pap. No. 4. 65 pp.

Wettenhall, R. L. 1975. Bushfire Disaster: An Australian Community in Crisis, Syd- ney, Aust: Argus, Robertson. 320 pp.

Wettenhall, R. L., Power, J. M. 1969. Bu- reaucracy and disaster I. Public Admin. 28:263-77

White, G., Haas, J. E. 1975. Assessment of Research on Natural Hazards, Cam- bridge, Mass: MIT Press. 230 pp.

Wilkinson, K. P., Ross, P. J. 1970. Citizen's Responses to Warnings ofHurricane Ca- mille. Mississippi State College, Miss:

Miss. State Univ., Soc. Sci. Res. Cent. Rep. No. 35. 60 pp.

Wolensky, R. P. 1975. The aftermath of the Great Agnes disaster: An analysis of emergent groups and local government officials in the Wyoming Valley of Penn- sylvania. PhD thesis. Penn. State Univ., University Park. 197 pp.

Wolf, C. 1975. Group perspective formation and strategies of identity in a post- threat situation. Sociol. Q. 16:401-14

Wright, J. E. 1976. Interorganizational systems and networks in mass casualty situations. PhD thesis. Ohio State Univ., Columbus. 117 pp.

Wright, J. E. 1977. Looting and Anti-Social Behavior in Disasters. Columbus: Ohio State Univ., Disaster Res. Cent. Rep. Ser. No. 20. 75 pp.

Zurcher, L. A. 1968. Social-psychological functions of ephemeral roles: A disaster work crew. Hum. Organ. 27:281-97



Response to Social Crisis and Disaster

Documents