| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||
COMMENTARY |
1 Department of Human Development, Cornell University, Ithaca, New York.
2 Department of Sociology, Purdue University, West Lafayette, Indiana.
Address correspondence to Karl Pillemer, PhD, Department of Human Development, Cornell University, Ithaca, NY 14853. E-mail: kap6{at}cornell.edu
IN recent years, both researchers and policy makers have turned their attention to relations between the generations after children reach adulthood. New demographic realities have spurred this trend: Increases in longevity have greatly lengthened the shared lifetimes of generations, and many people can look forward to continued relationships with their parents until well into the children's middle age. As research on intergenerational relationships has burgeoned across the social sciences, there has been concern that new conceptual approaches are needed. Given the rapidly changing and sometimes confusing nature of intergenerational relations in contemporary Western societies, it is probably not surprising that a common thread among emerging models is the concept of complexity.
This movement is both necessary and predictable in our field. In a classic article, the noted scientist and educator Warren Weaver described the key progression in the history of science as a movement from simpler models to more complex ones. In the early stages of a field, concern is with straightforward questions of categorization, description, and relatively simple hypotheses, whereas as the discipline progresses, the organized complexity of systems is acknowledged and investigated (Weaver, 1948
). It is clear that such a movement is under way in the scientific study of intergenerational relations among adults. Scholars from a variety of disciplines are moving beyond simple models to orientations and approaches that recognize the complex and sometimes contradictory world of parent–adult child relationships.
A recent article by Russell Ward (2008) is an outstanding example of how social scientists interested in the family in later life can take the complexity of relationships into account and shed new light on old questions. In this innovative and thought-provoking piece, Ward tackles two types of complexity. First, rather than taking a traditional unidirectional focus on closeness and attachment or tension and conflict, the article focuses on ambivalence—although, as we discuss shortly, not as usually defined. Second, instead of examining a single parent–child relationship, the research investigates how parents' relationships with multiple children within the same family differ. Ward further gives attention to gender differences throughout the analyses, an approach that is often recommended but not frequently accomplished, given that many studies focus solely on mothers.
For these reasons, we believe this article makes a substantial contribution to the literature. However, as is perhaps appropriate in this context, we find ourselves ambivalent about Ward's approach. On the one hand, we applaud the effort to take complexity into consideration, exploiting a well-established data set in creative new ways. On the other hand, the article raises questions regarding the operationalization and measurement of ambivalence that are somewhat perplexing. These issues are worthy of discussion and debate, and they alert us to definitional concerns about intergenerational ambivalence that need to be untangled by future investigations.
We find Ward to be entirely convincing in showing the benefits of obtaining reports on multiple children within the same family. He begins by noting an apparent contradiction in the study of older parent–adult child relations. It is clear family ties play an increasingly salient role as people grow older and that intergenerational relationships can be strong sources of support and fulfillment. However, there is less than convincing evidence from research thus far regarding the effect of relationships with children on parents' well-being. Ward argues that one way to shed light on this issue is to take into account "multiple relations" between parents and all of their children.
In contrast to most previous research, which has focused on a single parent–child dyad in each family or asked about adult children in the aggregate, Ward takes the critically important step of examining how parents differentiate among children in the same family. The clear payoff, as Ward notes, is a "more complete view of parent–child relations." By including all children, Ward is able to identify families in which there is at least one child with whom relationship quality is low. In what may be analogous to the maxim "You are only as happy as your unhappiest child," having one child in the sibship with whom the relationship is poor appears to trump other, more positive relationships. The innovative analysis in this article clearly provides critically important detail lacking in other research.
We encountered difficulties, however, in Ward's conceptualization and measurement of ambivalence. Ward introduces the concept of "collective ambivalence," for which there is, to our knowledge, no precedent in the literature. Although he does not provide a specific definition, from the discussion the term appears to designate a parent having different feelings regarding his or her individual children. Thus, if a parent has predominantly positive relationships with some children and predominantly negative ones with other children, he or she is identified in this model as "ambivalent." Although, as noted, we concur that looking at within-family variations in relationship quality is of great importance, it does not seem to us that Ward's approach fits any definition of ambivalence of which we are aware.
Attempts to define and measure ambivalence have a relatively long history. Despite major differences in the fields in which ambivalence has been studied, the definitions are surprisingly consistent in one respect. The most important single defining characteristic of ambivalence is a contradictory assessment of or response toward the same object. That is, virtually all treatments of ambivalence include simultaneous, contradictory attitudes or feelings as the core defining characteristic. In psychoanalysis, where the term originated, ambivalence is characterized in the extreme terms of simultaneous love and hate for a parent or child (Parker, 1995). In attitudinal research, ambivalence is defined as holding both positive and negative beliefs or feelings toward an object (Maio, Fincham, Regalia, & Paleari, 2004). Sociologists use the term to describe societal dilemmas or dichotomies in which social norms demand contradictory behaviors (such as autonomy and dependence) and people find themselves striving for two opposing goals simultaneously (Merton & Barber, 1963; Smelser, 1998).
In Ward's article, however, there is no measurement of simultaneous and contradictory attitudes and feelings of any kind. Rather, the ubiquitous phenomenon of differential relationships among children in the same family is referred to as ambivalence, stretching the concept to the point where it loses its meaning. In particular, we suggest that there is a unit of analysis problem here. If there is "collective ambivalence," then measurement needs to be about the collectivity, not the individual components. Otherwise, any situation in which an individual likes some, but not all, objects in a category would constitute ambivalence, which does not appear to us to be logical. It is possible that having good relationships with some children and poor ones with others may lead to simultaneous and contradictory attitudes or feelings toward something—family life, for example, or the concept of having adult children. But it is hard to see what, precisely, is ambivalent in any of the measures used in Ward's article. It is not clear why having different relations with various children would constitute ambivalence.
It may help to take the example of a teacher in a classroom, who (as probably all teachers do) likes some of his students and dislikes others. Is this teacher experiencing ambivalence? He is indeed doing so when he has both positive and negative feelings toward a particular student—for example, when he is angry that Jennifer, the class clown, is disruptive but simultaneously admires her originality and spirit. The resulting ambivalence may have effects on outcomes such as his sense of self-efficacy as a teacher or even his daily psychological well-being. However, does simply having some students the teacher likes and others he dislikes equal ambivalence? Such a conceptualization lacks any sense of being torn in two directions or experiencing simultaneously positive and negative emotions or cognitions toward a person or relationship.
Now the same teacher, because of his variously positive and negative relationships with multiple students, may feel ambivalent about the class as a whole, or about teaching as a career. One could even imagine him confiding to a friend, "I have kind of a love/hate relationship with that class." But in that case, one would have to develop measures about the larger entity. Otherwise, this line of thinking could stretch to absurdity, such as referring to someone who likes some foods and does not like others as being "ambivalent" about eating.
Furthermore, Ward's assessment of relationship quality does not lend itself to measuring ambivalence. In ambivalence research, the most widely accepted approach requires individuals to express opposing attitudes. For example, Fingerman and Hay (2004) asked individuals to identify parents or children both to whom they were emotionally close and who bothered them, and Pillemer and colleagues (2007) asked directly about "feeling torn" or "having mixed feelings" toward a child. Another established method is to ask respondents separately about the positive aspects of a relationship and about the negative ones; a mathematical algorithm is then used to calculate an ambivalence score (Willson, Shuey, & Elder, 2003). In contrast, Ward trichotomizes a single measure of closeness, actually contrasting good-quality relationships with less good-quality relationships. This measurement approach does not embody the contradictory feelings that the ambivalence perspective requires.
Because we feel so strongly that Ward's general approach is precisely the direction in which the field should move, we worry that our objections may appear to be nitpicking. However, as Ward notes, the study of ambivalence in intergenerational relations is growing rapidly, and this topic appears to be a very fertile area for research in the coming years. It is therefore important at precisely this stage in the development of the paradigm that conceptual clarity about ambivalence be achieved. In this case, not only is the general conceptualization flawed, but the actual items do not appear to measure ambivalence.
Thus, Ward's article raises the critically important issue of measurement. Developing a diversity of reliable and valid measures of intergenerational ambivalence is clearly necessary to advance scholarship on this topic. How should researchers approach the creation of such measures? We suggest that the most fruitful route is to build upon established measures in the field. One of the most crucial aspects of this process is systematically comparing new measures with those that have been used widely and are accepted by family scholars. This approach would be consistent with the standards used throughout social science in the development of new measures of phenomena that have already been subject to investigation, such as psychological well-being or marital quality. Building upon existing measures would ensure that the new measures would, in fact, allow scholars to shed new light on ambivalence among family members.
In conclusion, despite our differences regarding the nature and operationalization of ambivalence, Ward's approach is to be commended because it raises important questions about the future direction of research on intergenerational relations. The issues we have been discussing, and in particular within-family designs and the complexity of conceptualizing and measuring intergenerational relations, lead one to speculate about what the field will look like a decade from now. As noted, until recently researchers have relied almost exclusively on studies using between-family designs and have typically conceptualized and measured parent–adult child relations in a unidimensional manner focusing on traditional measures of closeness and conflict.
As a field, gerontology has learned a great deal from this body of work; in fact, this work has provided the conceptual and methodological foundation upon which the study of intergenerational relations now stands. But is one phase in this vigorous research area ending and another beginning? At this point, work that engages family complexity through conceptual and methodological approaches like Ward's is seen as "alternative" to the mainstream of the study of families in the middle and later years. However, as there is greater utilization of these designs and ideas, might they become the mainstream in another decade, much as longitudinal designs and the use of sophisticated multivariate models have become the norm across the past two decades?
Given the ability of these new approaches to answer questions that have not heretofore been raised among scholars but that resonate as important issues to both professionals who work directly with families and the general public, this change seems likely. Building on related recent work in parent–child relations earlier in the life course, we must ask the following question: Is it ever appropriate to generalize about parent–child relations from a "one parent, one child" study? More pointed, is it ever meaningful to use measures that ask for global assessments about relationships with "children in general"? Finally, is it now necessary to acknowledge that intergenerational relations involve a complex interplay of positive and negative feelings, attitudes, and behaviors? We understand that these questions mandate studies of intergenerational relations that are both complicated and labor intensive. This may be the necessary price to pay, however, for increasing our understanding of the complex world of parent–adult child relations in the 21st century.
Footnotes
Decision Editor: Kenneth F. Ferraro, PhD
Received for publication July 31, 2008. Accepted for publication August 4, 2008.
References
| ||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
