The Importance of Father Love
History and Contemporary Evidence

This article explores the cultural construction of fatherhood in America, as well as the consequences of this construction as a motivator for understudying fathers–especially father love–for nearly a century in developmental and family research. It then reviews evidence from 6 categories of empirical studies showing the powerful influence of fathers' love on children's and young adults' social, emotional, and cognitive development and functioning. Much of this evidence suggests that the influence of father love on offspring's development is as great as and occasionally greater than the influence of mother love. Some studies conclude that father love is the sole significant predictor of specific outcomes after controlling for the influence of mother love. Overall, father love appears to be as heavily implicated as mother love in offsprings' psychological well-being and health, as well as in an array of psychological and behavioral problems.

For most people, life's major satisfactions and pain revolve around personal relationships with others ( Duck, 1988 , 1991 ; Rohner, 1994 , 1999 ). For children, the most powerful of these others are parents. ¹ A vast literature shows that the quality of personal relationships–especially personal relationships with parents for children–is a major predictor of psychosocial functioning and development for both children and adults. One dramatically important component of the concept "quality of relationship" has to do with warmth, supportiveness, comforting, caring, nurturance, affection, or simply love. In the context of parent—child relationships, we summarize these elements under the construct parental acceptance—rejection or, more broadly, under the rubric of the warmth dimension of parenting ( Rohner, 1986 , 1999 ). Four decades of cross-cultural and intracultural research on issues of parental acceptance—rejection by Rohner (1960 , 1975 , 1986 , 2001 ) show that parents anywhere in the world can express their love or lack of love in any one or a combination of four major ways. Parents can, for example, be warm and affectionate (or cold and unaffectionate), hostile and aggressive, or indifferent and neglecting, or they can engage in undifferentiated rejection. Undifferentiated rejection refers to individuals' affectively charged belief that their parents do (or did) not really care about them, want them, or love them, without necessarily having clear behavioral indicators that the parents are (or were) unaffectionate, aggressive, or neglecting toward them. These and other such related concepts as parental support, nurturance, closeness, and caring–concepts that are often used more or less interchangeably by researchers–are all central elements in the overarching construct of parental acceptance—rejection or, simply, parental love.

Research in every major ethnic group of America ( Rohner, 2001 ), in dozens of nations internationally ( Khaleque & Rohner, in press ; Rohner & Britner, in press ), and with several hundred societies in two major cross-cultural comparative samples ( Rohner, 1975 , 1986 , 2001 ) has shown that children and adults everywhere–regardless of differences in race, language, gender, or culture–appear to respond in the same way when they experience themselves to be loved (accepted) or unloved (rejected) by the people most important to them growing up. The overwhelming bulk of research dealing with parental acceptance—rejection concentrates on mothers' behavior, however. Historically, the possible influence of fathers' behavior has been largely ignored.

This article discusses evidence regarding the relative sparseness of research on fathers, especially on father love. It then explores the cultural construction of fatherhood in America and the consequences of this construction as a principal motivator for overlooking fathers to a large degree for nearly a century of developmental studies. Finally, it discusses growing evidence about and implications of the recent recognition of the powerful influence of fathers' love in child development ( Rohner, 1998 ). Before continuing, we need to specify that this article concentrates on evidence regarding the influence of fathers' love-related behaviors–or simply father love–in relation to the social, emotional, and cognitive development and functioning of children, adolescents, and adult offspring. This emphasis on the importance of father love should not be construed as minimizing the well-documented importance of mother love. Rather, it is intended to emphasize the need to consider the influence of fathers as well as mothers whenever possible. Finally, where relevant, this article also addresses some of the antecedents of the effects of fathers' love, including gender and ethnicity. Other antecedents, including marital quality, separation and divorce, and children's temperament, for example, have been addressed elsewhere ( Booth & Crouter, 1998 ; Lupton & Barclay, 1997 ; Phares, 1996 , 1997 ).

Little is known about parents' actual behavior within American families before the 1930s, when empirical research on children and families had its fullest beginnings. Most of what is known about child rearing before that time comes from such sources as popular magazines, medical and religious books, journals, and biographies. These texts tended to exhort parents–almost always mothers–to behave in a particular way. Or they claimed that parents (mothers) behaved in a particular way without providing evidence that the claim was true. In addition, some authors made sweeping but undocumented generalizations about the effects of maternal (but rarely paternal) behavior. Some, for example, went so far as to place the entire burden of children's well-being in this life and the next on mothers' shoulders. In 1849 , for instance, Elizabeth Hall wrote in Mother's Assistant magazine:

Yes, mothers, in a certain sense, the destiny of a redeemed world is put into your hands; it is for you to say, whether your children shall be respectable and happy here and prepared for a glorious immortality, or whether they shall dishonor you, and perhaps bring your grey hairs in sorrow to the grave, and sink down themselves at last to eternal despair! (p. 27)

Earlier, in the 1700s, Rousseau had already proclaimed that "mothers love will cure society's ills" (as cited in Kagan, 1978 , p. 54).

For more than 200 years, mother love was generally considered paramount in child development ( Kagan, 1978 ; Stearns, 1991 ; Stendler, 1950 ; Sunley, 1955 ). Moreover, mothers were assumed to have nearly exclusive daily responsibility for the care of children. Fathers were seldom mentioned in the popular press before the mid-1920s ( Atkinson & Blackwelder, 1993 ), except occasionally as breadwinners, disciplinarians, teachers, and moral preceptors. Father love was virtually unrecognized by the media. A dramatic shift occurred with respect to gender-specific parenting articles in the mid-1920s, however. To illustrate, Atkinson and Blackwelder reviewed a sample of 1,482 popular magazine articles from 1900 to 1989. They found that by the mid-1920s gender-nonspecific "parenting" articles had begun to reach ascendancy over "mothering" articles. That is, the term parent began to supplant the term mother in most articles. It is quite possible, though, as noted by Atkinson and Blackwelder, that as women began entering the labor force in increasing numbers, popular writers found it more appropriate to use the term parenting when in fact they really meant mothering . In any case, only 16% of all articles published dealt explicitly with fathers. This percentage fluctuated little in the popular press throughout the course of the 20th century.

This evidence regarding the rarity of reports about fathers and especially about father love in the popular press is supported by Ellner (1973) , who reviewed every article related to child rearing for the first 6 months of 1950, 1960, and 1970 in three family monthly magazines ( Ladies Home Journal , Good Housekeeping , and Parents Magazine ). Of the 177 articles reviewed, only 3 dealt with fathers. In these, fathers were urged to participate in the discipline of their children and in children's sex education. Beyond that, fathers were viewed as important role models for sons and as husband models for daughters. None of the articles dealt with father love.

Early professional publications in child development and family studies reflect similar trends to those found in the popular press. Peterson, Becker, Hellmer, Shoemaker, and Quay (1959) , for example, reviewed the professional literature on parent—child relations from 1929 to 1956. They found about 169 publications dealing with mother—child relationships but only 12 (7%) dealing with father—child relationships. Eron, Banta, Walder, and Laulicht (1961) supported this conclusion when they estimated that about 15 times as many publications dealt with mother—child relations as with father—child relations. Furthermore, they found that when information was collected about fathers, it was commonly obtained from men's wives, not from fathers themselves. Nash (1965) argued that this tendency to obtain information about fathers from mothers (or from children) resulted from the researchers' implicit assumption that fathers themselves were inaccessible because of out-of-home economic responsibilities.

It is difficult to determine from these early reviews of professional literature what percentage of the few studies that dealt with fathers also implicated father love in some way. An estimate may be made, however, from the work of Rohner and Nielsen (1978) . These authors completed a critical review of the literature dealing with parental acceptance—rejection from about 1930 through 1976. In their work they found about 600 relevant studies. Of these, 108 (17%) mentioned fathers, but 70% of the latter references occurred in the 1960s and 1970s, a time when many other researchers were also recognizing fathers as significant parenting figures. Phares and Compas (1992) , for example, reviewed every article in eight clinical and child and adolescent journals from 1984 through 1991. The authors wanted to ascertain possible gender bias in the reporting of parental influences in child and adolescent psychopathology. They found that 48% of the articles reviewed included only mothers, whereas 1% included only fathers. However, 26% of the studies obtained and analyzed data separately for both mothers and fathers. These figures are representative of the fact that a virtual revolution has occurred since the 1960s and 1970s in the recognition of fathers, albeit only minor recognition of the influence of father love ( Biller, 1974 , 1981 , 1993 ; Hanson & Bozett, 1985 , 1991 ; Hewlett, 1992 ; Lamb, 1975 , 1981 , 1986 , 1997 ; Mackey, 1996 ; Radin, 1981 ). We turn now to an exploration of the following question: Why has father love been so understudied for nearly a century in research on parent—child relations?

Though usually unintended and often unrecognized, much of behavioral science is a value-laden enterprise ( Kaplan, 1964 ; Silverstein & Auerbach, 1999 ). Research questions that are regarded as appropriate or sensible at a particular point in time are usually situated within a matrix of cultural beliefs often widely accepted within the dominant population at large, but certainly within the scientific community more specifically. The issue of fatherhood is a case in point. Fatherhood is a cultural construction ( Doherty, Kouneski, & Erickson, 1998 ), and once formulated it has implications for the subsequent behavior of those who share the beliefs and assumptions defining that construction.

The issue here is to understand the meanings commonly associated in the United States with the concepts of father, fatherhood, and fathering and to recognize the constraints that are implicitly placed on one's behaviors as a result of accepting these meanings (i.e., cultural construction) as being true. Of course, to understand fully the cultural construction of fatherhood, one must also understand its counterpoint, motherhood. And one must also understand that both constructions are influenced by cultural conceptions of masculinity and femininity. It is not our intention here, however, to disentangle the maze of meanings and behavioral implications of all of these conceptions and relationships, but merely to acknowledge that they exist. For now we want only to dwell on the cultural meaning of fatherhood in America in relation to motherhood and to explore the implications of this conception for behavioral science research on parent—child relationships.

For most Americans, the concepts father, fatherhood, and fathering appear to connote very different domains of behavior and affect from the concepts mother, motherhood, and mothering. Semantically, the cultural construction of "fathering," for example, implies nothing about father love as does the genderized equivalent of mother love that is contained semantically in the construct "mothering." The word mothering elicits, for many, a warm, fuzzy, nurtured feeling, whereas the term fathering elicits feelings of something stronger, colder, harder, and less affectionate ( Rohner, 1995 ). The term father love is not used in everyday discourse, but mother love is. Even the phrase "father love" sounds strange to some, yet many feel comforted by "mother love" ( Rohner, 1995 ). Popular literature is filled with references to both mothering and parenting when referring to caregiving. But the term fathering is almost never used in this context. When used, the term is typically found in the context of the question "Who begot whom?" Even the gender-neutral term parent is often used or interpreted as being synonymous with mother . And we have found in our own teaching that students–including advanced graduate students–often misread and even mishear the term paternal as being parental , which is then sometimes translated as maternal/mother .

In some respects, the conception of fatherhood has shifted dramatically over the course of the last 300 years of American history ( Lamb, 2000 ). According to E. H. Pleck and Pleck (1997) , for example, the ideal image of the colonial European American father in the 1600s and 1700s was that of the stern patriarch. From 1830 to 1900 the ideal image of father was one of the distant breadwinner. From 1900 to 1970 the ideal father was the genial playmate dad and gender role model. And from the 1970s to today the ideal image of father is said to be one who is a co-parent, sharing equally with his mate in the care of their children. This portrayal of the cultural conception of the ideal father is greatly oversimplified, of course, but it does contain important ideas that have been widely shared, and it shows how cultural conceptions of fatherhood have shifted over time.

One of the most enduring historical elements defining fatherhood has nothing to do with rearing children but deals with the assumption that the major role of fathers within the family is as economic provider: the breadwinner ( La Rossa, 1997 ). Throughout much of the 20th century and earlier–beginning with the advent of industrialization in the 19th century–many American men judged themselves (and were judged by others), judged their personal worth, and judged their success as husbands and fathers in relation to their ability to provide economically for their families ( Stearns, 1991 ). Since the 1940s, this essential role has been deemphasized somewhat, although it has not been altogether abandoned or replaced. The shared ideology of male breadwinning was used by many researchers over the course of the 20th century to explain fathers' apparently limited involvement in child care ( Griswold, 1993 ).

More important, however, many behavioral scientists prior to the 1960s and 1970s assumed that fathers were relatively unimportant for the healthy development of their children in any case ( Cabrera, Tamis-LeMonda, Bradley, Hofferth, & Lamb, 2000 ; Nash, 1965 ; Rapoport, Rapoport, Strelitz, & Kew, 1977 ). At the very most, fathers were thought to be peripheral to the job of parenting because children spent most of their time with their mothers. Some even argued that fathers have no biological aptitude for child care, though women were said to be genetically endowed for it ( Amato, 1998 ; Belsky, 1998 ; Benson, 1968 ). These conceptions seemed to have led in the popular press and in many television portrayals during the 1950s and 1960s to an image of the irrelevant, mindless, ineffectual, sometimes bumbling and incompetent father ( Mackey, 1996 ). Not all literary and visual portrayals of fathers were negative, however. For example, movies and television series such as Make Room for Daddy, Father Knows Bes, The Cosby Show , and others portrayed warm, loving, involved, and competent fathers. Similarly, some 19th- and 20th-century novels described loving, concerned, and competent fathers. Bob Cratchit in Dickens's (1843) A Christmas Carol and Atticus Finch in Lee's (1960) To Kill a Mockingbird are two cases in point. And even though comic strip fathers were sometimes lampooned in the 20th century as being bumbling incompetents, La Rossa, Jaret, Gadgil, and Wynn (2000) also showed that these characters were frequently portrayed from 1940 to 1999 as being nurturant, supportive, and capable fathers.

The cultural conception of fathers as being a relatively inconsequential influence in child development has its counterpoint in an assumption about the primacy of mothers and mother love. We commented earlier on the latter issue. Now we amplify on the topic by pointing out that virtually all of the influential theories of child development during the past century accepted the unproven premise that mothers are most important in child development. For example, psychoanalytic theory specifically and psychodynamic theories more generally assume that the mother—child relationship–especially during the first 6 years of life–is crucial for normal child development. Early attachment theory too focused on the mother—child bond, as did significant portions of learning theory and cognitive developmental theory ( Lamb, 1975 , 1981 , 1986 ; Phares, 1992, 1996 ). With this phalanx of distinguished theories converging on a view that was widely believed in America anyway, few developmental researchers in the first half of the 20th century thought it essential to inquire directly about fathers' influence in child development. Shared wisdom asserted that competent and nurturant mothering was all children really needed for successful cognitive, social, and emotional development ( Bronstein & Cowan, 1988 ). Or, as one distinguished researcher wrote in an anonymous review of this article: "We spent a lot of time studying mothers because we thought they were important, not because we thought fathers weren't."

Cultural assumptions such as these about fathering are related to widely shared assumptions about appropriate gender role and gender identity of both men and women. In this regard, most Americans–men as well as women–seem to have associated child care with feminine behavior. Until a decade or so ago, many Americans regarded it as unmasculine for men to spend much time caring for children, except as a temporary backup and support to the mothers. A common theoretical assumption found even as recently as the 1970s and 1980s asserted that fathers' major contribution to child development was indirect, through their economic and emotional support of mothers ( Biller, 1993 ; Maccoby & Martin, 1983 ). This premise may be summarized as follows: The most important contribution a father can make to his children is achieved through providing for and loving their mother. In summary, the widely held cultural construction of fatherhood prior to the 1970s (and still held by many today) has two strands. One deals with fathers' competence as caregivers, and the other deals with the influence of fathering on child development. The first strand asserts that fathers are often incompetent and maybe even biologically unsuited to the job of child rearing. (The maternal counterpoint is that women are genetically endowed for child care.) The second strand asserts that fathers' influence in child development is relatively unimportant or at least peripheral or indirect. (The maternal counterpoint is that mother love and competent maternal care provide everything that children need for normal, healthy development.)

The effect of internalizing these cultural beliefs as one's own personal beliefs seems to have led to sometimes unintended and unrecognized but nonetheless real consequences. The most notable outcome of acting on these beliefs throughout most of the 20th century was to minimize fathers' presence in much of mainstream behavioral science research as well as in clinical research ( Phares, 1997 ). Because it was assumed that mothers but not necessarily fathers were important in child development, researchers tended to study mostly mothers' behavior. Of course, they found significant effects of maternal behavior that served to motivate researchers even further to study mothers. But a subtle side effect of these results also seemed to reinforce researchers' belief that fathers must not be all that important because mothers were being shown to be so very important. Moreover, behavioral scientists and clinical practitioners felt further justified in excluding fathers from their work because mothers spent the greatest amount of time with children, and therefore it was reasoned that they must also have the greatest influence on children's development as well as on treatment outcomes ( Ferholt & Gurwitt, 1982 ; Phares, 1996 ). Finally, fathers tended to be omitted from research and treatment models because they were assumed to be inaccessible as a result of their out-of-home economic responsibilities.

The relatively little research that included fathers prior to 1970 dealt with a variety of issues, but seldom with father love per se. This omission seems to have been encouraged by the widespread "tendency among both researchers and theorists to accept without question the assumption that fathers express less affection and understanding toward children than do mothers" ( Walters & Stinnett, 1971 , p. 102). Indeed, many behavioral scientists and clinicians seemed to accept the postulate that fathers' major role in the family was in the instrumental domain, whereas mothers' major role was in the expressive—affective domain ( Parsons & Bales, 1955 ). Acceptance of these assumptions resulted in very limited research specifically examining the relationship between father love and child outcomes.

Though it was not until the 1960s that researchers began to find with any regularity that father love was as predictive as mother love of children's psychological and behavioral adjustment, occasional note was made of this fact as early as the 1940s. R. W. Lidz and Lidz (1949) , for example, claimed that faulty paternal influences were as common as maternal influences in the development of child psychopathology. Later, T. Lidz, Parker, and Cornelison (1956) claimed that domineering, sadistic, and rejecting fathers were more implicated in the etiology of schizophrenia than were mothers. Peterson et al. (1959) completed one of the first studies examining the attitudes of both fathers and mothers and their effects on both disturbed and normal children. From this research, the authors noted that "contrary to general assumption and our own original expectation, the attitudes of fathers were found to be at least as intimately related as the attitudes of mothers to the occurrence and form of maladjustive tendencies among children" (p. 129).

Empirical evidence such as this about the strong influence of father love led Becker, Peterson, Hellmer, Shoemaker, and Quay (1959) and Becker (1960) to make an emphatic call for more systematic study of the role of fathers in child development. And later, in his classic review of the literature on parent—child relations, Becker (1964) wrote, "where both mothers and fathers have been studied, most of the research has shown the father's influence on the child's behavior to be at least equal to that of the mother" (p. 204). For the most part, this early evidence and call for additional research about the possible influence of fathers' behavior was ignored because the then current cultural ideology continued to endorse the primacy-of-the-mother doctrine.

Nearly a decade later, in a major review of the 1960s parent—child relationship literature, Walters and Stinnett (1971) reiterated these observations:

Because we have believed that the impact of mothers upon the development of children is greater than the impact of fathers, we have investigated maternal impact to a far greater extent than we have examined the impact of fathers. Yet, much of the evidence of the past decade suggests that the variability of children's behavior is more closely associated with the type of father one has than the type of mother [italics added]. (p. 102)

Other early sources of evidence about the importance of fathering and father love also went more or less unheeded because of the cultural bias in America emphasizing the singular importance of mothers. For example, anthropologists have documented repeatedly the fact that women the world over tend to be the major caregivers of children, but they are not the exclusive caregivers ( Mackey, 1996 ; Rohner & Rohner, 1981 ). Fathers, siblings, grandparents, and others are common substitutes. As early as 1956 the noted anthropologist, Margaret Mead, concluded that "anthropological evidence gives no support...to the value of such an accentuation of the tie between mother and child. On the contrary, cross-cultural studies suggest that adjustment is most facilitated if the child is cared for by many warm, friendly people" (pp. 642—643). Fathers often rank among the most significant of these others.

In the same general domain, results from Rohner's (1975) early research on the worldwide antecedents and correlates of parental acceptance—rejection also went mostly unheeded. More specifically, in a cross-cultural comparative study of 101 societies representing a stratified sample of the known and adequately described sociocultural systems of the world, Rohner found that children everywhere–across the full range of the world's economic systems, political systems, household types, and other sociocultural factors–tend to be accepted by mothers and other major caregivers to a greater degree in households where fathers are present on a day-to-day basis than in households where fathers are present less often. Moreover, results of this research showed that the more important fathers are as socializing agents in relation to other caregivers such as mothers, siblings, and grandparents, the greater the warmth children receive from all major caregivers in that society. In addition, the more time fathers willingly spend tending their offspring in relation to other caregivers, the more likely children are to be accepted. These results were replicated but essentially ignored in a subsequent holocultural study of 186 societies worldwide ( Rohner, 1986 ; Rohner & Rohner, 1982 ).

This interest in gender role redefinition was fueled by the fact that mothers with children were entering the out-of-home workforce in unprecedented numbers. In 1960, for example, fewer than 40% of American mothers worked out of the home, but by 1972 almost 50% were doing so ( Biller, 1993 ). By 1997, 68% of the mothers in the United States with children less than 18 years old were engaged in full-time or part-time employment ( Child Trends, 2000 ). As greater numbers of mothers began working, the feminist movement called for fathers to become more involved in child care and household tasks ( Biller, 1993 ; Bronstein & Cowan, 1988 ; Griswold, 1993 ). To be successfully accomplished, these changes required a partial redefinition of traditional male and female gender roles ( Jain, Belsky & Crnic, 1996 ; Mackey, 1996 ). This in turn brought the role of fathering and the influence of father love into sharp research focus ( Maccoby & Martin, 1983 ). For the first time, some behavioral scientists became keenly motivated to study fathers directly ( Bronstein & Cowan, 1988 ; Giveans & Robinson, 1985 ; Parke, 1985 ).

It is not our purpose here to try to fully disentangle the snarled web of competing social, political, economic, and cultural forces associated with the feminist movement that eventually politicized fatherhood in America ( Griswold, 1993 ). Here we simply want to recognize that as a direct as well as indirect result of the feminist movement, many behavioral scientists began to study fathers and father love directly. And when they did, they found that fathers are as capable as mothers of being competent and nurturant caregivers ( Bronstein & Cowan, 1988 ; Silverstein & Auerbach, 1999 ). They also found that the father—child bond often parallels the mother—child bond both emotionally and in intensity ( Fox, Kimmerly, & Schafer, 1991 ; Hanson & Bozett, 1991 ).

A number of the studies reviewed subsequently continued to show, as in earlier decades, that fathers' influence was as great as mothers' for specific developmental outcomes. But it was not until the 1990s that the third source of influence came into full play, causing many behavioral scientists to fully recognize that fathers should be included in studies dealing with parent—child relations. This source of influence derived from research results based on readily accessible, easily used, and powerful multivariate statistical packages. Use of multivariate statistics, including multiple regression and structural equation modeling (SEM), allowed investigators to control simultaneously for the influence of a variety of variables. In doing so, researchers discovered that father love sometimes explains a unique, independent portion of the variance in specific child outcomes, over and above the portion of variance explained by mother love. Indeed, some studies reviewed later found that father love is the sole significant predictor of specific child outcomes after removing the influence of mother love.

Six categories of empirical studies show the influence of father love on specific child outcomes: (a) Some studies look exclusively at the influence of variations of father love without examining the influence of mother love too; (b) some conclude that father love is equally as important as mother love in predicting specific child outcomes; (c) some conclude that father love predicts specific child outcomes better than mother love; (d) some conclude that father love is the sole significant predictor of specific child outcomes after removing the influence of mother love; (e) some conclude that father love moderates ( Baron & Kenny, 1986 ) the influence of mother love on specific child outcomes; and, finally, (f) some conclude that paternal versus maternal parenting may be associated with a single outcome or with different outcomes in sons and daughters. These six categories emerged naturalistically from the growing body of empirical research dealing with the influence of father love. Three of the categories (the second through fourth) are especially noteworthy because they show that father love continues to make a unique and significant contribution to child outcomes after statistically controling for the influence of mother love.

We briefly review each of these six categories of evidence. Of course, many other aspects of fathers' behavior, some of which may be closely related to father love–such as father involvement, father absence, and fathers' psychological and behavioral state–also influence child development. Here, though, we focus on the impact of father love (or paternal acceptance—rejection) per se because, as studies reviewed subsequently show, father love by itself is implicated in a wide array of developmental issues. These include youths' psychological adjustment, behavior problems, delinquency, gender role development, cognitive/academic/intellectual development, achievement, and social competence. Moreover, father love has also been shown to be associated with children's and adults' psychological health and sense of well-being. Even though some of these studies conclude that father love is the sole significant predictor of specific outcomes–after controlling for the influence of mother love–none suggest that mother love is unimportant in other contexts. We should note here that authors of the articles reviewed subsequently use a variety of terms to discuss different aspects of the father love (paternal acceptance—rejection) construct. Many of these concepts, such as paternal warmth, nurturance, support, caring, and affection, are used more or less interchangeably and synonymously with paternal acceptance—rejection. Because of this, we generally retain the major terms used by the various authors. Sample characteristics of the empirical studies cited in this article are provided in Table 1 .

Many studies looking exclusively at the influence of variations in father love deal with two topics: (a) gender role development ( Fish & Biller, 1973 ; Green, 1982 ; Huttenen, 1992 ; Millen & Roll, 1997 ; Musser & Fleck, 1983 ) and (b) father involvement ( Amato & Rivera, 1999 ; Biller, 1993 ; K. N. Harris, Furstenberg, & Marmer, 1998 ; Radin, 1981 ). Studies of gender role development emerged prominently in the 1940s and continued through the 1970s ( Lamb, 1997 , 2000 ). This was a time when fathers were considered to be especially important as gender role models for sons. Commonly, researchers assessed the masculinity of fathers and of sons and then correlated the two sets of scores. Many behavioral scientists were surprised to discover that no consistent results emerged from this research until they examined the quality of the father—son relationship. Then they found that when the relationship between masculine fathers and their sons was warm and loving, the boys were indeed more masculine. Later, however, researchers found that the masculinity of fathers per se did not seem to make much difference after all. ² As summarized by Lamb (1997) :

Boys seemed to conform to the sex-role standards of their culture when their relationships with their fathers were warm, regardless of how "masculine" the fathers were, even though warmth and intimacy have traditionally been seen as feminine characteristics. A similar conclusion was suggested by research on other aspects of psychosocial adjustment and on achievement: Paternal warmth or closeness appeared beneficial, whereas paternal masculinity appeared irrelevant. (p. 9)

The second domain in which a substantial amount of research has been done on the influence of variations in father love deals with father involvement, that is, with the amount of time that fathers spend with their children (engagement), the extent to which fathers make themselves available to their children (accessibility), and the extent to which they take responsibility for their children's care and welfare (responsibility; Lamb, Pleck, Chernov, & Levine, 1987 ). Many studies conclude that children with highly involved fathers, in relation to children with less involved fathers, tend to be more cognitively and socially competent, less inclined toward gender stereotyping, more empathic, and psychologically better adjusted ( Biller, 1981 , 1993 ; Easterbrooks & Goldberg, 1984 ; Lamb, 1997 ; J. H. Pleck, 1997 ; Radin, 1981 ; Radin & Russell, 1983 ; Radin & Sagi, 1982 ; Radin, Williams, & Coggins, 1993 ; Reuter & Biller, 1973 ; E. Williams & Radin, 1993 ; S. Williams & Finley, 1997 ). Commonly, these studies investigate both paternal warmth and paternal involvement and find–using simple correlations–that the two variables are related to each other and to youth outcomes.

It is unclear from these studies whether involvement and warmth make independent or joint contributions to youth outcomes. Moreover, "caring for" children is not necessarily the same thing as "caring about" them. Indeed, Lamb (1997) concluded from his review of studies of paternal involvement that it was not the simple fact of paternal engagement (i.e., direct interaction with the child), availability, or responsibility for child care that was associated with these outcomes. Rather, it appears that the quality of the father—child relationship made the greatest difference ( Cabrera et al., 2000 ; Lamb, 1997 ). J. H. Pleck (1997) reiterated this conclusion when he wrote:

The critical question is: How good is the evidence that fathers' amount of involvement, without taking into account its content and quality, is consequential for children, mothers, or fathers themselves? The associations with desirable outcomes found in much research are actually with positive forms of paternal involvement, not involvement per se. Involvement needs to be combined with qualitative dimensions of paternal behavior through the concept of "positive paternal involvement" developed here. (pp. 66—67)

Research by Veneziano and Rohner (1998) supports these conclusions. In a biracial sample of 63 African American and European American children, the authors found from multiple regression analyses that father involvement by itself was associated with children's psychological adjustment primarily insofar as it was perceived by youths to be an expression of paternal warmth (acceptance). These results varied by ethnicity, however. In the European American families, fathers' loving acceptance significantly mediated ( Baron & Kenny, 1986 ) the way offspring experienced their fathers' involvement. In the African American families, however, father involvement made no significant contribution to youths' psychological adjustment, but perceived paternal warmth (acceptance) did. Neither parenting nor youths' psychological adjustment varied significantly, however, by social class.

Further evidence in support of Pleck's conclusion about the importance of paternal warmth versus levels of paternal involvement is shown in Veneziano's (1998) cross-cultural comparative, holocultural study. Using multiple regression analysis, in a sample of 32 societies representing the world's known and adequately described sociocultural systems, he found that the lack of paternal warmth–not the amount of time that fathers were involved with children– predicted young males' interpersonal violence.

Finally, Amato and Gilbreth's (1999) meta-analysis of 63 studies that explored the relationship between nonresidential fathering and children's well-being showed that paternal encouragement, support, and closeness were more predictive of youths' social, emotional, and psychological well-being than the frequency of contact between children and their nonresident fathers.

As we indicated earlier, many of the studies concluding that father love is as influential as mother love go back to the 1940s. Most of these conclusions, especially those prior to the 1980s, are drawn from correlational studies in which the simple correlation between a specific measure of paternal love and a specific child outcome is as great as or greater than the simple correlation between the same measure of maternal love and the child outcome. More recently, however, the 1980s and 1990s saw behavioral scientists use forms of multivariate analyses that allowed them to conclude that both fathers' and mothers' behaviors are associated significantly and uniquely with specific child outcomes.

The great majority of studies showing that father love is as important as mother love deal with one or a combination of the following five issues among children, adolescents, and young adults: (a) personality and psychological adjustment problems including issues of self-concept/self-esteem, emotional stability, and aggression ( Amato, 1998 ; Becker, 1960 ; Becker et al., 1959 ; Buri, 1989 ; Buri, Louiselle, Misukanis, & Mueller, 1988 ; Buri, Murphy, Richtsmeier, & Komar, 1992 ; Dekovic & Meeus, 1997 ; Emmelkamp & Karsdorp, 1987 ; Fine, Voydanoff, & Donnelly, 1993 ; Jacobs, Spilken, & Norman, 1972 ; McPherson, 1974 ; Monkman, 1958 ; Nash, 1965 ; Peppin, 1962 ; Peterson et al., 1959 ; Peterson, Becker, Shoemaker, Luria, & Hellmer, 1961 ; Sears, 1970 ; Yamasaki, 1990 ); (b) conduct problems, especially in school ( Becker, 1960 ; DeKlyen, Biernbaum, Speltz, & Greenberg, 1998 ; DeKlyen, Speltz, & Greenberg, 1998 ; McPherson, 1974 ; Paley, Conger, & Harold, 2000 ; Patterson, Reid, & Dishion, 1992 ; Renk, Phares, & Epps, 1999 ; Russell & Russell, 1996 ; Siantz & Smith, 1994 ); (c) cognitive and academic performance issues ( Amato, 1998 ; Carroll, 1973 ; Easterbrooks & Goldberg, 1984 ; Heilbrun, Orr, & Harrell, 1966 ; Peppin, 1962 ); (d) mental illness ( Arrindell, Emmelkamp, Monsma, & Brilman, 1983 ; Crook, Raskin, & Eliot, 1981 ; Greenberger & Chen, 1996 ; R. W. Lidz & Lidz, 1949 ; T. Lidz et al., 1956 ; Richter, Richter, & Eisemann, 1990 ); and (e) substance abuse ( Barnes, 1984 ; Prendergast & Schaefer, 1974 ).

Only newer studies in the 1980s and 1990s used complex multivariate analyses to systematically test for the relative influence of fathers' love in contrast to mothers' love. The work of Young, Miller, Norton, and Hill (1995) illustrates this. These authors drew from a national sample of 640 adolescents 12 to 16 years old living in two-parent families. Employing SEM techniques, they found that perceived paternal love and caring were as predictive of sons' and daughters' life satisfaction–including their sense of well-being–as maternal love and caring. Review of a broad range of studies such as these led Lamb (1997) to conclude that

fathers and mothers seem to influence their children in similar rather than dissimilar ways. Contrary to the expectations of many psychologists, including myself, who have studied paternal influences on children, the differences between mothers and fathers appear much less important than the similarities. Students of socialization have consistently found that parental warmth, nurturance, and closeness are associated with positive child outcomes whether the parent or adult involved is a mother or a father. The important dimensions of parental influence are those that have to do with parental characteristics rather than gender-related characteristics. (p. 13)

Two types of studies are common in this category. First, results of some bivariate correlational studies have led researchers to conclude that fathers' love is more strongly associated than mothers' love with specific child behaviors such as those noted subsequently. Second, the 1980s and especially the 1990s saw a proliferation of studies using multiple regression and SEM. As these analytic procedures became more commonplace, it also became more common to discover that the influence of father love explains a unique, independent portion of the variance in specific child outcomes–detailed subsequently–over and above the portion of variance explained by mother love.

Studies in this category tend to deal with one or more of the following six issues among children, adolescents, and young adults: (a) personality and psychological adjustment problems ( Amato, 1994 ; Dominy, Johnson, & Koch, 2000 ; Komarovsky, 1976 ; Stagner, 1933 ; Tacon & Caldera, 2001 ), (b) conduct problems ( Chen, Liu, & Li, 2000 ; Eron et al., 1961 ; Grant et al., 2000 ), (c) delinquency ( Andry, 1962 ), (d) mental illness ( Barrera & Garrison-Jones, 1992 ; Lefkowitz & Tesiny, 1984 ), (e) substance abuse ( Brook & Brook, 1988 ; Emmelkamp & Heeres, 1988 ), and (f) psychological health and well-being ( Amato, 1994 ). We briefly review an example of each issue.

Research by Dominy et al. (2000) illustrates the first issue. In this work, the authors studied the relation between perceived childhood experiences of parental (maternal and paternal) acceptance—rejection and the presence or absence of binge eating disorder (BED) among a group of 113 women. Eighty-three of these women were obese. Of these, 32 suffered from BED. Thirty women, however, neither were obese nor had an eating disorder. The authors found that women with BED perceived their fathers to be significantly more rejecting than did women in either of the other two groups. In addition, the women with BED perceived their fathers to be significantly more rejecting than their mothers.

A 2-year longitudinal study of 258 sixth-grade students (at Time 1) in the People's Republic of China serves to illustrate the second issue: dealing with conduct problems. In this study Chen et al. (2000) found that paternal but not maternal warmth at Time 1 was negatively associated with youths' disruptive aggression toward peers at Time 2. In addition, they found that paternal but not maternal warmth was positively associated with children's academic performance and social competence as judged by teachers.

An older work by Andry (1962) illustrates the third type of study: the link between paternal and maternal rejection and delinquency. Here Andry found in a matched sample of 80 delinquent boys 11 through 15 years of age that the great majority of delinquents felt rejected by their fathers but not necessarily by their mothers. Nondelinquents, on the other hand, tended to feel loved by both parents. Of special interest in Andry's study is the fact that fathers as well as mothers in the two groups tended to corroborate the youths' perceptions of parenting.

Research by Rohner and Brothers (1999) illustrates the fourth issue, dealing with the link between parenting and mental illness. These authors examined the relation between perceived parental (paternal and maternal) acceptance—rejection and self-reported psychological adjustment in a sample of 17 women diagnosed as suffering from borderline personality disorder. This group was compared with a control sample of 18 nonclinical women. Results showed that women in the clinical group tended to perceive qualitatively more paternal rejection than acceptance in their families of origin. Perceived maternal rejection was also elevated somewhat, but not to the point where these women experienced qualitatively more maternal rejection than acceptance. And, as is generally true throughout much of America ( Rohner, 1986 , 2001 ), most of the women in the nonclinical control group tended to perceive substantial paternal as well as maternal love and overall acceptance.

Research by Campo and Rohner (1992) illustrates the fifth issue, dealing with the relation between parenting and substance abuse. These researchers studied 40 drug abusers in relation to a control sample of 40 nonabusers. On the average, the polydrug-addicted group–both males and females–had experienced qualitatively more paternal rejection than acceptance before the onset of serious drug use as adolescents. The substance abusers had also experienced significant love withdrawal at the hands of their mothers, but not to the point of having experienced qualitatively more rejection than acceptance. Participants in the nonabuse group, on the other hand, had experienced substantial paternal as well as maternal love and acceptance in their families of origin. Discriminant function analysis showed that perceived paternal acceptance—rejection, self-reported psychological adjustment, perceived maternal acceptance—rejection, and level of education (in that order of importance) predicted with 91.2% accuracy who among the 80 participants were drug abusers versus nonabusers. Perceptions of father love and love withdrawal were overwhelmingly the best single predictor.

Finally, Amato's (1994) work serves to illustrate the sixth issue. Here, Amato showed that father love is sometimes implicated to a greater extent than mother love in adult offsprings' overall psychological health and well-being. Drawing from telephone interviews with a national sample of 471 young adults, he found that perceived closeness to fathers for both sons and daughters made a unique contribution–over and above the contribution made by perceived closeness to mothers–to adult offsprings' happiness, life satisfaction, and low psychological distress (i.e., overall psychological well-being). In the author's words, "Regardless of the quality of the mother-child relationship, the closer adult offspring were to their fathers, the happier, more satisfied, and less distressed they reported being" (p. 1039).

A growing number of studies using a variety of multivariate statistics have begun to conclude that father love is occasionally the sole significant predictor of specific child outcomes after removing the influence of mother love. Studies in this category tend to deal most often with one or more of the following three issues among children, adolescents, and young adults: (a) personality and psychological adjustment problems ( Barnett, Marshall, & Pleck, 1992 ; Bartle, Anderson, & Sabatelli, 1989 ; Dickie et al., 1997 ; DuBos, Eitel, & Felner, 1994 ; Matsuda & Ritblatt, 1998 ; Wagner & Philips, 1992 ), (b) conduct and delinquency problems ( Kroupa, 1988 ), and (c) substance abuse ( Brook, Whiteman, & Gordon, 1981 ; Eldred, Brown, & Mahabir, 1974 ).

Cole and McPherson (1993) , for example, concluded from their SEM analysis of 107 adolescents and parents that father—child conflict but not mother—child conflict (each controlling for the other) was positively associated with adolescent depressive symptoms. Moreover, father—adolescent cohesion was positively associated with the absence of adolescent depressive symptoms. These results are consistent with Barrera and Garrison-Jones's (1992) conclusion that adolescents' satisfaction with paternal support was related to adolescent depression, whereas maternal support was not. A similar study by Barnett et al. (1992) was conducted among 285 married adult sons with two living parents. When measures of the quality of both mother—son and father—son relationships were entered simultaneously into a regression equation, only the father—son relationship was significantly related to adult sons' psychological distress (a summed measure of anxiety and depression).

In a sample of 70 middle- to lower-middle-class adolescents, Forehand and Nousiainen (1993) assessed relations between school functioning, as evaluated by teachers, and youths' perceptions of parental (paternal and maternal) acceptance—rejection. Using stepwise multiple regression, the authors found that teachers' ratings of youths' social competence and conduct problems were associated with the adolescents' perceptions of paternal but not maternal acceptance.

A 6-year longitudinal study begun in 1981 by Brody, Moore, and Glei (1994) showed that paternal (but not maternal) warmth had a significant long-term effect in shaping adolescents' attitudes in 1987 toward such social issues as marriage, divorce, sex roles, child support, welfare, and teenage childbearing. More specifically, the warmer fathers were and the more adolescents were allowed to participate in family decision making in 1981, the more adolescents internalized their parents' values over time as being their own. This relationship was true only for warm father—adolescent relationships. Warm mother—adolescent relationships had no significant effect on youths' later attitudes. Results from this study were based on 592 families participating in a nationally representative household survey of 11- to 16-year-old youths.

Finally, Veneziano (2000) found in a sample of 281 African American and European American families that only paternal warmth was significantly related to the European American youths' psychological adjustment when controlling for the influence of maternal warmth. Indeed, maternal warmth dropped from the regression model altogether. However, in the African American families, paternal as well as maternal warmth was significantly related to youths' psychological adjustment, making both independent and joint contributions.

Fathers' behavior may moderate and be moderated by–that is, interact with ( Baron & Kenny, 1986 ; Hull, Tedlie, & Zahn, 1992 )–other influences within the family. Apparently, however, only one study so far has addressed the issue of mother love having different effects on specific child outcomes depending on the level of father love. Here, Forehand and Nousiainen (1993) found in a sample of 70 adolescents and their parents that the relative level of fathers' love and acceptance moderated the contribution of mothers' loving acceptance to adolescent functioning. Specifically, variation in father love contributed to adolescents' cognitive competence as well as to anxiety—withdrawal (internalizing problems) through its interaction with maternal acceptance. Regarding cognitive competence, for example, when mothers were high in acceptance but fathers were low, teachers judged youths' cognitive competence to be quite low, lower even than when mothers' acceptance was also low. But when fathers' loving acceptance was high, mothers' loving acceptance was associated with the most positive levels of cognitive competence. From this study, the authors drew the important inference that "simply including fathers in parent-adolescent research is insufficient. Instead, the potential ways in which each parent's style contributes to the other parent's style must be considered" (p. 219).

Two types of research tend to be found in this category. First, some research shows that one pattern of paternal behavior and a different pattern of maternal behavior is associated with a single outcome in sons, daughters, or sometimes both offspring. Second, other research in this category shows that a single pattern of paternal love-related behavior is associated with one outcome for sons and a different outcome for daughters. We briefly review examples of both types.

The work of Barber and Thomas (1986) illustrates the first type. These authors found in a sample of 527 adolescents that daughters' self-esteem was best predicted by fathers' physical affection (kisses and hugs) and by mothers' general support, including maternal praise, approval, encouragement, use of terms of endearment, and helping behaviors. Sons' self-esteem, on the other hand, was best predicted by fathers' sustained contact (e.g., picking up the boy for fun and safety) and by mothers' companionship (i.e., spending time with the boy and sharing activities with him). In addition, in perhaps the earliest study ever done of this type, Fitz-Simmons (1935) found that paternal rejection and maternal overprotection characterized the parenting styles among 96 emotionally withdrawn children being treated in child guidance clinics. Moreover, a substantial literature finds differences in paternal versus maternal love-related behaviors influencing different aspects of the gender role development of both boys and girls ( Biller & Borstelmann, 1967 ; Bronson, 1959 ; Distler, 1965 ; Kelly & Worell, 1976 ; Mussen, 1961 ; Mussen & Distler, 1959 ; Orlofsky, 1979 ; Payne & Mussen, 1956 ).

The work of Jordan, Radin, and Epstein (1975) serves to illustrate the second type of study in this category, that is, studies concluding that a single pattern of paternal love-related behavior is associated with different outcomes for sons versus daughters. These authors found in a sample of 180 European American children 4 years of age that paternal nurturance was positively associated with middle-class sons' performance on the Stanford Binet Intelligence Scale ( Terman & Merrill, 1960 ). Paternal nurturance was unrelated, however, to daughters' performance on the scale.

Finally, a second study is also broadly illustrative of this second type of research. Here Booth and Amato (1994) found in a longitudinal study of 419 parents and their adult children that a poor relationship between spouses while children were living at home was associated 12 years later with somewhat different outcomes for sons and daughters. Specifically, adult sons felt somewhat less close to both parents than did sons whose parents had had a good marital relationship. Daughters, on the other hand, felt much less close to their fathers but only slightly less close to their mothers. Thus, the authors concluded, the father—daughter tie tends to be especially vulnerable in the context of serious marital problems between parents, whereas the mother—daughter tie tends to be especially resilient.

Several problems and limitations characterize this body of research. First, most of the work appears to deal with middle-class European American parents. Research on the influence of father love among African Americans, Hispanic Americans, Asian Americans, and Native Americans is relatively uncommon, and it is equally rare in cross-cultural studies. As a result, it remains for future research to determine the extent to which results reported here can be generalized beyond middle-class European American families. Second, many of the studies reviewed here relied on a single source (e.g., youths' or parents' reports) to provide information about both parental acceptance—rejection and the relevant outcome variable(s). As noted by Campbell and Fiske (1959) ; Marsiglio, Amato, Day, and Lamb (2000) ; and others, however, shared-method variance in studies such as these might artificially inflate the correlation between variables, resulting in an overestimate of the true effect size.

Third, even though it seems unmistakably clear that father love makes an important contribution to offsprings' development and psychological functioning, it is not at all clear why paternal acceptance—rejection is sometimes more strongly associated with specific child outcomes than is maternal acceptance—rejection. And it is unclear why patterns of paternal versus maternal parenting are sometimes associated with different outcomes for sons, daughters, or both children. Part of the reason for these differences no doubt lies in the fact that fathers and mothers often interact with their children in somewhat different ways ( Fagot, 1995 ). Fathers generally interact with children less frequently than do mothers, for example, and they generally tend to be less involved in caregiving ( Fagot, 1995 ; K. N. Harris et al., 1998 ). Moreover, when they do interact with their children, fathers often initiate different types of behaviors from those of mothers. For example, fathers tend to engage in more physical, rough-and-tumble, and idiosyncratic play than do mothers ( Collins & Russell, 1991 ; Forehand & Nousiainen, 1993 ; Parke, 1996 ). In addition, fathers are more likely than mothers to encourage children's competitiveness and independence and to encourage their children to take risks ( Cabrera et al., 2000 ). Nothing in these behaviors per se, however, appears to explain why father love sometimes has different effects from mother love.

Possibly, evolutionary and behavior-genetic influences should be considered in attempts to explain some of these effects ( Collins, Maccoby, Steinberg, Hetherington, & Bornstein, 2000 , 2001 ; J. R. Harris, 1998 ; Rowe, 1994 , 2001 ). Some of these differences, however, are probably related to the fact that fathers are often perceived to have more power and authority within the family than do mothers ( Radin, 1981 ). It remains for future research to inquire directly about the relevance of these mechanisms. Until then we can know that father love is often as influential as mother love–and sometimes more so–but we cannot know for sure why this is true.

Proper recognition of the influence of fathers and father love should have several salutary effects. First, widespread recognition of fathers' influence may help motivate many men to become more involved in nurturing child care. Simons, Whitbeck, Conger, and Melby (1990) supported this point when they found that "fathers are more likely to engage in nurturing activities when they believe such behaviors will make an important difference in the life of their child" (p. 387). Of course, the degree to which men are likely to become involved in nurturing paternal behavior is at least partly a function of their mates' internalized cultural beliefs about the maternal role ( Parke, 1995 ). That is, because of the enormous cultural emphasis placed on the role of motherhood in America, some women are ambivalent about encouraging fathers to become heavily involved as nurturing parents ( Allen & Hawkins, 1999 ; Biller, 1993 ; Cowan & Bronstein, 1988 ; Doherty et al., 1998 ; Marsiglio et al., 2000 ). And of course, as already mentioned, some men have difficulty redefining masculinity to include nurturing paternal behavior. Nonetheless, the evidence seems clear that mothers are more effective parents when fathers are both supportive partners and nurturing parents. And children are major beneficiaries when both parents are warm and loving.

A second salutary effect of widespread recognition of the influence of fathers and their love should be to help reduce the incidence of "mother blaming" common in the clinical field. That is, according to Phares (1997) and others ( Caplan, 1986 , 1989 ; Caplan & Hall-McCorquodale, 1985 ; Phares & Compas, 1992 ), the great emphasis on mothers and mothering in clinical research and practice–without examining the influence of fathers and fathering–has led to a tendency toward blaming mothers for children's maladjustment and psychopathology.

The emphasis in this review on fathers should not be construed to suggest that mother love is generally less important than father love. Indeed, some studies discussing the effects of father love relative to mother love conclude that mother love has greater influence than father love for specific developmental outcomes. For example, in their meta-analysis of 15 studies dealing with the relation between gender of parent and youths' externalizing behaviors, Rothbaum and Weisz (1994) found that the correlation between caregiving and externalizing was larger–but apparently not statistically so–for mothers than for fathers in 80% of the studies reviewed. According to a related meta-analysis of the relation between caregiving and conduct problems—delinquency by Loeber and Stouthamer-Loeber (1986) , however, 64% of the 75 studies reviewed reported that fathers' love-related behaviors had a stronger association with these child outcomes than did mothers' behavior.

Evidence such as this punctuates the need to include both fathers (and other significant males, when appropriate) and mothers in future research and then to analyze the data for possible father and mother effects separately ( K. N. Harris et al., 1998 ). It is only by separating data in this way that behavioral scientists can discern when and under what conditions paternal and maternal factors have similar or different effects on specific child outcomes ( Cole & McPherson, 1993 ; Kim & Rohner, 2001 ; Phares, 1996 ; Phares & Compas, 1992 ). This call for separate measurement and analyses of maternal and paternal influences is contrary to the argument made by some ( Kurdek & Fine, 1994 ; Kurdek & Sinclair, 1988 ; Schwartz, Barton-Henry, & Pruzinsky, 1985 ) that behavioral scientists should combine mothers' behavior and fathers' behavior into a single, composite index.

Beyond this, evidence cited in this article makes even more compelling the call by Silverstein and Phares (1996) and others (e.g., Jessor, 1993 ; Kerr & Bowen, 1988 ; Parke, 1995 ; Phares, 1996 ) to "transform the dominant theoretical paradigm in the social sciences from a dyadic [i.e., mother—child] to a triadic [i.e., father—mother—child], or larger systems model" ( Silverstein & Phares, 1996 , p. 48). This transformation would include a paradigm shift in guidelines for conducting research as well as in graduate training and clinical programs to automatically include fathers as well as mothers in all parenting matters. In addition, this call recognizes the need to explore social policy implications of research showing the powerful influence of father love as well as mother love.

Finally, we should note that many questions remain unresolved and even unexplored in the research literature assessing the relative contributions of mothers' and fathers' love and love withdrawal. For example, little is known about the question of whether children might be differentially affected by mother love in comparison with father love at different ages or at different developmental stages. Similarly, it is unclear whether the magnitude of the effect of mother love or father love varies across offsprings' life span. And, as already mentioned, little is yet known about why the influence of father love is sometimes greater than the influence of mother love. Answers to these and many other such questions await future research.

1

In this article, the term parents is defined as whoever the most important caregivers are of a child. These people are not necessarily mothers or fathers. But typically they are.

It would be misleading to imply that all studies of boys' gender role development dealt with father—son relationships only. The mother—son relationship was sometimes included. Overall, however, results suggested that a nurturant father—son relationship is more important to a boy's development of masculinity than is the mother—son relationship ( Biller & Borstelmann, 1967 ).