The Relationship Between Smile Type and Play Type ... - CiteSeerX

13 downloads 333 Views 1003KB Size Report
vocational school or at least some college classes. Procedure ...... behavior of infants and their dual-career parents. ... Orlando, FL: Academic Press. Fogel, A.
Developmental Psychology 1997, Vol. 33, No. 6, 925-933

Copyright 1997 by the American Psychological Association, Inc. 0012-1649/97/$3.00

The Relationship Between Smile Type and Play Type During Parent-Infant Play K. Laurie Dickson

Heather Walker and Alan Fogel

Northern Arizona University

University of Utah

This study examined the relationship between smile type and play type during parent-infant interactions in the home. Thirty-six mother-infant and father-infant dyads were videotaped playing for 10 min. Smile type (basic, Duchenne, and duplay smiles) and play type (object, physical, vocal, and book reading) were coded. Results of loglinear analysis indicated that different types of smiles occur during different types of play more often than expected if distributed equally. In addition, different smile-type and play-type patterns occurred for father-infant dyads compared with mother-infant dyads. Qualitative analyses were used to generate hypotheses about the reasons why different types of smiles occurred during various play activities.

Previous research regarding parent-infant play and infant smiling has lumped together different types of smiles and primarily examined differences in frequency of smiling directed toward mothers and fathers or only has examined mothers. Several researchers have shown that infants tend to smile more during interactions with their mothers (Ban & Lewis, 1974; Feldman & Ingham, 1975; Field, Vega-Lahr, Goldstein, & Scafidi, 1987). Yet others have found that infants often show a preference for their fathers in their display of affective behaviors, such as smiling, laughing, looking, and offering toys (Belsky, Gilstrap, & Rovine, 1984; Clarke-Stewart, 1978; Lamb, 1976a, 1976b, 1977). These differences could be due to parental play style, as research reveals that mothers and fathers have different styles of play with their young children. Fathers have a greater propensity for play of a highly physical and idiosyncratic nature; that is, they tend to be more tactile and arousing with their infants than do mothers (Lamb, 1977; Power & Parke, 1982; \bgman, 1982). Mother-infant play is more likely to involve visual games, object play, and conventional social games, such as peekaboo (Clarke-Stewart, 1978; "togman, 1982).

in the psychological and ethological literature for humans and nonhuman primates (Dedo, 1991; Ekman, Davidson, & Friesen, 1990; Fox & Davidson, 1988; Messinger, Fogel, & Dickson, 1997; van Hooff, 1972). Three specific facial actions are typically assessed when studying smile types; lip corner raise, cheek raise, and jaw drop or mouth openness. Given that each of these facial actions are independent, various combinations of these facial actions are possible. This study examined four different smile types. A basic smile involves a lip corner raise that pulls the lips upward and toward the side of the face. In play smiles, the jaw drops, and the lip corners raise. With a Duchenne smile, the lip corners and the cheeks raise, which crinkles the eye corners of adults (Ekman & Friesen, 1978). A duplay smile includes a lip corner raise, a jaw drop, and a cheek raise. Researchers disagree on both the meaning of these different types of smiles and the reasons why they exist. For instance, some investigators (Andrew, 1972; Fridlund, 1991; Kraut, 1978; Smith, 1985) argued that facial expressions are social interactive signals or communications of intent, yet others contend that they are reflections of varying emotions or varying degrees of emotion intensity (Demos, 1982; Ekman & Friesen, 1982). Although this study was not designed to assess either the meaning of these different smiles or the reasons why they exist, it may provide information that will shed light on these questions. Different types of smiles reliably occur in different social contexts and correlate with specific patterns of brain electrical activation (Blurton-Jones, 1972; Cheyne, 1976; Dedo, 1991; Ekman & Friesen, 1982; Ekman, Davidson, & Friesen, 1990; Fox & Davidson, 1988; Jones, Raag, & Collins, 1990; Messinger, 1994; Messinger et al., 1997). Duchenne smiles are more often associated with pleasant stimuli and self-reports of pleasure in adults than are basic smiles (Ekman et al., 1990). In 10-month-old infants, Duchenne smiles are associated with mother's approach, whereas basic smiles are associated with stranger's approach (Fox & Davidson, 1988). For infants and toddlers, open-mouth smiles are more common while playing with an attentive partner, whereas closed-mouth smiles occur more frequently during toy-centered play (Cheyne, 1976; Demos, 1982; Jones & Raag, 1989). Jones et al. (1990) also found

In the current study, attention is paid to the relationship between morphologically different smiles and various types of play during parent-infant interactions in the home. The focus is not on parent gender differences per se; rather, the focus is on the types of smiles that occur during different types of play during mother-infant and father-infant interactions. Morphologically different types of smiles have been identified

K. Laurie Dickson, Department of Psychology, Northern Arizona University; Heather Walker and Alan Fbgel, Department of Psychology, University of Utah. The article is based on K. Laurie Dickson's doctoral dissertation at the University of Utah. We would like to thank JoNell Strough and Katerina Calderone for their consultation on the statistical aspects of this project and the families that participated in the study. Correspondence concerning this article should be addressed to K. Laurie Dickson, Department of Psychology, Northern Arizona University, Flagstaff, Arizona 86011. Electronic mail may be sent via Internet to [email protected]. 925

926

DICKSON, WALKER, AND FOGEL

that 17-month-old infants' bared-teeth smiles were more often directed toward the mother than an object; thus, they were more likely to occur in social versus nonsocial contexts. BlurtonJones (1972) reported a reliable association between three mouth positions and social context. First, during social exchanges, children most often displayed smiles with lips parted and teeth showing, similar to a basic smile. Second, play-face smiles were associated with conditions of high excitement. Finally, smiles with lips together tended to occur during solitary activities. In this study, we used a unique combination of analyses, loglinear analysis and a descriptive narrative approach. Loglinear modeling examines higher order interactions between multilevel categorical variables and localizes the source of such interactions. Thus, patterns of actions that occur together can be revealed. Co-occurrences allow for the examination of organized action patterns as they occur over time. For example, if an infant looks at her father and smiles at the same time, this pattern would be different than if she looks at her father, then looks away and smiles. Temporal issues can be addressed with co-occurrences, whereas correlations do not afford this type of interpretation. Knowing that two behaviors occur at the same time is very different than knowing there is a correlation between two behaviors. The descriptive narratives provided an opportunity to identify aspects of the interaction that were not assessed in the a priori coding schemes but that may have contributed to the smileplay patterns. This qualitative approach also helped guide the interpretations of the loglinear findings by providing valuable observational information. This mode of analysis afforded a comprehensive assessment by characterizing the parent-infant play in a more detailed manner to capture the continuous nature of the ongoing interaction. More specifically, the qualitative analyses allowed for the generation of empirically driven "hypotheses that can be systematically tested in future studies. Twelve-month-old infants were observed in the home. Infants at this age are beginning to use words and therefore are becoming better participants in give-and-take play interactions. Other studies have examined both play behavior and positive facial expressions at approximately this age as well (i.e., Dedo, 1991; Fox & Davidson, 1988; Lamb, 1977; MacDonald & Parke, 1986; Stevenson, Leavitt, Thompson, & Roach, 1988). Little is known about the specific conditions under which infant smiles occur during play interactions in the home. Do different types of smiles occur during specific types of play? Are there different smile-type/play-type patterns for mother-infant and father-infant dyads? On the basis of the research questions and the decision that loglinear modeling was the most appropriate type of analysis, the hypotheses are stated in a manner consistent with the use of loglinear analysis. A two-way interaction of play type and smile type was predicted in that different types of smiles should occur more often than expected if distributed equally across cells during specific types of play. Previous research has found a relationship between basic smiles and object play (Cheyne, 1976; Demos, 1982; Jones & Raag, 1989). Given the similarities between object play and book reading, such as visual attention and the balance between cognitive and social stimulation, it is hypothesized that basic smiles will more likely occur during

both of these play types. Play smiles have been found to cooccur during rough-and-tumble style play for both humans and nonhuman primates (Blurton-Jones, 1972; van Hooff, 1972). It is unclear, given previous research, what the relationship will be between Duchenne smiles and play type. This study also examined differences in the smile-type/playtype relationship for mother-infant and father-infant dyads. Researchers have found that mothers and fathers play differently with children and that within specific play types, mothers and fathers interact differently (Clarke-Stewart, 1978; Lamb, 1977; MacDonald & Parke, 1984; Yogman, 1982). Given these findings and the evidence that different smile types occur during different play styles (Cheyne, 1976; Demos, 1982; Jones & Raag, 1989), a three-way interaction of smile type, play type, and parent gender was predicted. Specific effects were not predicted, given the lack of previous research regarding the relationship between parent gender, play type, and infant smile type. Walker (1994), using the same data set as in the current study, found no infant gender effects or birth order effects in relation to affect or play type. Walker's null findings led to the decision to forgo analyses of infant gender in relation to smile type and play type. Method

Participants Families were located through newspaper birth announcements in the Salt Lake City area. Letters were sent out to 400 potential families. The majority of interested families did not fit the requirement of the study (two children or less) or were unable to be located. Of the eligible families, approximately 5% refused to participate. Forty-three intact families with either firstborn or second-born 12-month-old infants were videotaped in the family home. Videotaping difficulties such as no audio (2 participants), poor picture quality (2 participants), and multiple breaks in the videotape (3 participants) forced the removal of 7 of the original 43 families that were videotaped. The participants included in the analyses consisted of 36 families with 17 female infants and 19 male infants (19 firstborn and 17 second born). The mean age for mothers was 27 years and was 30 years for fathers. The median household income was $30,000; the range was from $11,000 to $135,000. Sixty-six percent of the families were members of the Church of Jesus Christ of Latter-Day Saints. Approximately 49% of the mothers did not work outside the home, and 36% worked more than 20 hr a week outside the home. The average level of education for mothers and fathers was approximately the same; all had either attended vocational school or at least some college classes.

Procedure Each family home was visited by a graduate student and an undergraduate research assistant or two graduate students. Approximately 10 min were spent talking to the family prior to the videotaping in order for the infant and parents to feel comfortable with the observers, during which time the parents filled out the consent forms. Parents were instructed to ' 'play with their infant as they normally do" in the area of their choosing. Seventy-eight percent played in the living room, 6% played in the bedroom, and 16% played outside in the yard. Parents also decided which toys, if any, they would use during the play session. Each parent-infant dyad was videotaped in play for 10 min, without the presence of other family members. When possible, the observations were counterbalanced in terms of who played first; various reasons dictated the decision of

927

SMILE TYPE AND PLAY TYPE who played with the infant first (getting home late, shyness of one parent, etc.). Each parent filled out a demographic questionnaire while the other parent was being videotaped. Two videocameras were used to tape the play sessions. One camera provided an overall view of the play interactions (interaction tape), and the other camera provided a close-up view of the infant's face (infantface tape). Time base was laid on all videotaped sessions. An event at the beginning of the recorded session, such as the infant grabbing a book, was identified on the infant-face tape, and the time was noted. The exact same event was found on the interaction tape, and the time was noted. The official start time of the session corresponded with the identified event that indicated 0 s. The sessions were converted to this timetable by a trained research assistant. The start and end times from the infant-face videotapes were matched with the start and end times on the interaction videotapes. This procedure synchronized the time scales; therefore, the time associated with behaviors coded from one videotape corresponded to the behaviors that were coded from the other videotape.

Coding Smile and play behaviors were assessed continually in time on separate passes through the videotapes by trained coders. For each videotaped session, a trained coder viewed the tape until a defined change occurred. For example, if the father and infant were reading a book at the start of the play session, the coder would note the start time and the book reading code. When the dyad shifted from book reading to object play, the coder noted the precise time (in minutes and seconds) that the play code changed. This process occurred every time the dyad switched play type until the end of the session. A similar procedure was used for the smiletype coding. Play type. Two undergraduates in psychology were trained on the play-type coding scheme with pilot videotapes. The parent-infant interactions were categorized into the following play types. Object play was coded when the partners were playing with or using an object in a social manner; for instance, the parent uses a doll to kiss the infant, or the partners use a toy together. Object play was also coded when one partner was playing with an object, and the other partner was watching. Physical play was coded when the partners were in physical contact, such as rough-and-tumble play (wrestling or horsie), or during physical games such as tickling, "gonna get you," and pretend biting. Vocal play consisted of games such as singing along or peekaboo. Book reading was coded when the partners were looking at a book together. Nonengaged was coded when one of the partners was not involved with the other partner; for example, the infant was staring at the camera. Smile type. A coder certified in Ekman and Friesen's (1978) Facial Action Coding System (FACS) and trained on the infant version, Baby FACS (Oster & Rosenstein, in press), coded infant smiles from the videotape that focused on the infant's face. Facial expressions were coded according to action units (AU) that describe the specific muscle groups that are responsible for changing the facial features. In this study, the main features recorded include lip corner raise (AU12), cheek raise (AU6), and mouth openness (AU26/AU27). Lip corner raises were coded continuously, given that all types of smiles involve lip corner raises. For every lip corner raise, cheek raises and jaw drops were assessed. Various combinations of these facial actions are possible; this study assessed four smile types. Four smile types were classified depending on the combination of the facial action units. Basic smiles involve infraorbital triangle raising, nasolabial furrow deepening, and lip corner raising (AU12) caused by movement of the zygomatic major muscle. Duchenne smiles involve lip corner raises and cheek raises (AU6) caused by the contraction of the orbicularis oculi. Play smiles involve lip corner raises and a jaw drop due to movement of the pterygoid muscle (AU26/AU27). Duplay smiles consist of lip corner raising and both a jaw drop and a cheek raise.

Because of the nature of naturalistic observational data, can't see codes were required when either the infant's face was out of view of the camera or the dyad was out of view of the camera. Both videotapes were consulted before this code was used. Though precautions were taken by using two videocameras, unfortunately this procedure could not prevent all possible instances of obstruction. Frequencies of the coding categories are displayed in Table 1. Given the limitation of the modeling procedures and the focus on engaged play, both can't see codes and nonengaged codes were excluded from the analysis. Given the low frequency of play smiles relative to the other three smile types, the play smile category was eliminated from the analysis. A possible reason for the low frequency of play smiles is proposed in the Discussion section.

Reliability Interrater reliability was calculated on a random sample of 15% of the sessions for the play-type coding. Cohen's kappas, which measure agreement (correcting for random agreement) of the duration of all codes within a particular coding category, were calculated with a MicroAnalytic Data Analysis Package (see Kienapple, 1987). The mean kappa for the play coding was .95 (.91-.98 range). The first author coded 100% of the sessions, using the smile-type coding scheme. Another certified FACS coder viewed 21% of the sessions to assess reliability. The mean kappa for the smile coding was .82 (.63-.94 range).

Results The hypotheses were related to the following research questions: Do different types of smiles co-occur during different types of play? Do different smile-type/play-type patterns occur for mother-infant dyads and father-infant dyads? To avoid redundancy, all hypotheses were addressed with one analysis. A loglinear model was developed with smile type (basic, Duchenne, duplay), play type (object, physical, vocal, book reading), and parent gender (mother, father) to examine patterns of co-occurrences. Models containing the three-way interaction and various combinations of two-way interactions were compared against the data as to their ability to generate expected frequencies that were not significantly different from observed frequencies. The

Table 1 Frequency of Smile-Type and Play-Type Codes Category and code Smile type No smile Basic (AU12) Duchenne (AU12 and AU6) Play (AU12 and AU26/27) Duplay (AU12 and AU6 and AU26/27) Can't see Play type Object Physical Vocal Book reading Nonengaged Can't see Note.

AU = action unit.

Frequency

%

5,176 763 820 124 798 2,884

48.7 7.2 7.7 1.2 7.5 27.1

6,047 2,992 285 286 763 192

56.8 28.1 2.7 2.7 7.2 1.8

928

DICKSON, WALKER, AND FOGEL

goal of loglinear modeling is to develop a model based on linear equations that fit most parsimoniously with the data that reveals significant effects that are readily interpretable (Hopkins et al., 1990). As predicted, all the terms in the selected model were necessary to fit the model to the observed frequencies. The model produced three first-order effects: smile type, play type, and parent gender; two two-way associations: Parent Gender X Play Type and Smile Type X Play Type; and one three-way association: Smile Type X Play Type X Parent Gender. The analyses are based on events, such as smile type and play type, not on individual dyads. The number of cases (3,558) in relation to the number of cells (24) fit the requirement that the number of cases exceed five times the number of cells (Demaris, 1992; Green, 1988; Kennedy, 1992; Knoke & Burke, 1980; Tabachnick & Fidell, 1989). The loglinear model had adequate fit between observed and expected frequencies, likelihood ratio, G 2 (2) = 1.15 p = .563. Standardized parameter estimates, which resemble z scores (Green, 1988), were used to identify significant effects, that is, to localize differences in levels of variables from what would be expected by chance (Kennedy, 1992). The reported differences reflect significance at the .01 level; thus, the standardized parameter estimates were above 2.58. Given that the research questions concerned the issue of context specificity of smiles, the discussion focuses on interactions that were observed more often than expected if the data were distributed equally across the cells; therefore, the standardized parameter estimates are in the positive direction (see Tables 2 - 4 for the standardized parameter estimates for each interaction in the loglinear model). Although standardized parameter estimates provide information of the overall importance of an association, the effects of the co-occurrences between the variables are reported in terms of the observed percentage of events cross-classified into particular categories (Adams, LabouvieVief, Hobart, & Dorsosz, 1990; Tisak, 1986; Tisak & Turiel, 1984).

Type x Smile Type association indicated that duplay smiles occurred during physical and object play, whereas basic smiles occurred during book reading more often than expected, if distributed equally across the cells. The Smile Type X Play Type X Parent Gender association (see Table 4) illustrated that the smile-play relationship was different for mother-infant dyads and father—infant dyads. During mother-infant object play, Duchenne smiles occurred more often than expected, whereas during father-infant object play, basic smiles occurred more often than expected, if distributed equally across the cells. During mother-infant physical play, basic smiles occurred more often and duplay smiles occurred more often than would be expected, if distributed equally during father-infant physical play. During mother-infant book reading, duplay smiles occurred more often. However, during fatherinfant book reading, Duchenne smiles occurred more often than expected if distributed equally across the cells. The results of the quantitative analyses indicate in which context different types of smiles are more likely to occur; however, they provide little insight as to why. For instance, why were play smiles more likely to occur during physical play? Why were duplay smiles more likely to occur during object play? Why were basic smiles more likely to occur during book reading? Why might the relationship between smile type and play type be different for mother-infant and father-infant interactions? Qualitative analyses provided an opportunity to examine how aspects of the interaction that were not assessed in the a priori coding schemes contributed to the smile-play patterns. With both loglinear analysis and descriptive narratives, context specificity of different types of smiles during the parent-infant play was assessed at both macro and micro levels of description. By this means, empirically driven hypotheses were derived and can be systematically tested in future studies.

Qualitative Analyses

The main effects for smile type and play type indicated that object (65% of all play was object play) and physical (32.6%) play occurred more often than expected if the play types were distributed equally across the cells. Both basic (33.8%) and Duchenne (32.6%) smiles occurred more often than expected if the smile types were distributed equally. The two-way effects are shown in Tables 2 and 3. The Play Type x Parent Gender association revealed that mothers engaged in book reading, whereas fathers engaged in physical play more often than would be expected, if distributed equally. The Play

The positive significant effects from the three-way association (Smile Type X Play Type X Parent Gender) in the loglinear model were elaborated by creating descriptive narratives of select play sequences (see Appendix). To be selected for description, the infant's face was required to be in view of the camera for the entire sequence (between 15 and 20 s), and the infant had to smile at some point during the sequence, regardless of specific smile type. The play sequences were chosen randomly from the 45 episodes that fit the criteria. The resulting 6 descrip-

Table 2 Percentages and Standardized Parameter Estimates for Play Type X Parent Gender Play type Object

Physical

Vocal

Book reading

Parent gender

%

SPE

%

SPE

%

SPE

%

SPE

Mother Father

49.4 50.6

+.44 -.44

41.4 58.6

-7.90 +7.90*

58.9 41.1

+.83 -.83

73.5 26.5

+3.05* -3.05

Note. Asterisk indicates that the standardized parameter estimate is significant at the .01 level in the positive direction (>2.58). SPE = standardized parameter estimate.

929

SMILE TYPE AND PLAY TYPE

Table 3 Percentages and Standardized Parameter Estimates for Play Type X Smile Type Play type Object

Book reading

Vocal

Physical

Smile type

%

SPE

%

SPE

%

SPE

%

SPE

Basic Duchenne Duplay

38.8 33.7 27.6

-.22 -4.17 +2.86*

25.0 28.6 48.9

-5.37 -5.51 +7.38*

20.7 55.2 24.1

-4.11 +2.08 + 1.98

59.5 36.9 3.6

+5.45* +2.55 -4.68

Note. Asterisk indicates that the standardized parameter estimate is significant at the .01 level in the positive direction (>2.58). SPE = standardized parameter estimate.

Second, the qualitative analyses revealed that smiles typically occur in clusters and often do not occur as a single communicative behavior. A smile may start as a basic smile and transform into a Duchenne smile prior to becoming a duplay smile, all within one play sequence, in a manner that appears to correspond with actions in the social context. These clusters or runs have been found in the first 6 months of life as well (Fbgel, 1977, 1985; Kaye & Fbgel, 1980). Out of the 30 instances of smiling observed in the descriptive narratives, all but 4 were observed to occur in clusters. The descriptive narratives provide evidence of this dynamic flux of emerging smile types, which is suppressed when the social context is categorized into broad play categories and smile types are not delineated. Careful examination of the narratives revealed that certain localized actions or components of the interaction may help explain the emergence of specific smile types. First, the qualitative data suggest that basic smiles occur during visually demanding situations. Half of the 12 basic smiles that were observed in the descriptive narratives occurred when the infant was looking at the parent's face; the other half occurred when the infant was examining either an object or a picture. A sequence from the father-infant book reading narratives demonstrates this point:

tive narratives included a play sequence from both motherinfant and father-infant dyads for physical play, object play, and book reading. Given that the purpose of the qualitative analyses was to help characterize the parent-infant play in a more detailed manner than was possible with the a priori coding scheme, 6 narratives were deemed sufficient. Once the play sequence was selected, the first author repeatedly viewed the play sequences in both real time and slow motion to describe in detail each partner's contribution to the interaction. After the 6 narratives were completed, commonalities were extracted. These narratives allowed for the generation of empirically driven hypotheses that can be systematically tested in future studies and that was the primary goal of the qualitative analyses. The qualitative portion of the analyses produced two major findings. First, the findings from the loglinear analysis simplify the complex relationship between different types of smiles and different types of play. The loglinear findings demonstrate that smile types tend to be context specific. For example, duplay smiles are more likely than expected to occur during physical play, if distributed equally. However, the descriptive narratives reveal that, although different types of smiles predominate during certain types of play, all of the different smile types emerge during the different types of play as well. Thus, an unequivocal interpretation of the loglinear results is difficult because multiple types of smiles occur during the various play types.

After gazing at the camera person for four seconds, she bobs her head and Duchenne smiles as she looks toward the camera again

Table 4 Percentages and Standardized Parameter Estimates for Play Type X Smile Type X Parent Gender Play type Parent gender and smile type Mother Basic Duchenne Duplay Father Basic Duchenne Duplay

Physical

Object

Vocal

Book reading

%

SPE

%

SPE

%

SPE

%

SPE

35.7 37.9 26.9

-3.19 +4.71* -1.63

32.5 27.6 40.0

+4.87 -1.45 -3.82

18.4 61.8 19.7

-.12 +2.16 -1.49

60.7 33.9 5.4

-.8 -2.89 +2.74*

43.1 27.8 29.2

+3.19* -4.71 + 1.63

20.3 29.1 50.5

-4.87 + 1.45 +3.82*

25.0 42.5 32.5

+.12 -2.16 1.49

57.1 42.9 0

+.8 +2.89* -2.74

Note. Asterisk indicates that the corresponding standardized parameter estimate is significant at the .01 level in the positive direction (>2.58). SPE = standardized parameter estimate.

930

DICKSON, WALKER, AND FOGEL

and giggles. Her cheeks lower just prior to her lowering her head to look at the pictures with a basic smile. As she raises her head to look in the same direction as she had seconds before, her cheeks raise again. Then, the infant Duchenne smiles as the father turns the page and she raises her head to look back at the camera. Her cheeks lower into a basic smile as her gaze returns to the pictures and raise again as she looks back up at the camera. Second, the descriptive narratives also suggest that tactile stimulation occurs simultaneously with duplay or play smiles. Out of the eight duplay-play smiles (one was a play smile) observed in the descriptive narratives, all except two occurred during tactile stimulation. Both play and duplay smiles were assessed in the narratives, and no attempt was made to differentiate them in the interpretation, given the low frequency of play smiles. For example, during the father-infant physical play sequence, As she rolls, the father lowers his head into her stomach, saying "I'm gonna get you." Her cheeks raise and her jaw drops into a duplay smile as his face touches her stomach and she begins to laugh. Her vocalizationsfluctuateat the same speed in which he is tickling her stomach with his face and fingers in a side to side fashion. The infant's jaw closes into a basic smile while she looks at him, as the father withdraws his physical stimulation. The smile is gone as she looks away and her vocalizations change in character from laughing to a minor distress call.

Discussion The purpose of this study was to assess the relationship between different types of smiles and parent-infant play interactions. The descriptive narratives allowed for examination of components of the interaction that may influence the occurrence of different types of smiles during play. The primary goal was to document the current findings within the literature and speculate as to possible reasons for the specific smile-type/play-type patterns that were revealed in the data.

Relationship Between Smile Type and Play Type Previous research (Dedo, 1991) found that, with mothers, basic smiles occurred more during object play than during direct face-to-face interactions. This study found that basic smiles occurred during book reading. The present finding is similar in that attention during book reading is not focused on the mother directly as in face-to-face interactions; rather, attention is focused on a book, which may be similar to attention being focused on an object. Basic smiles, which involve no cheek raises, may occur during book reading because of the primary focus on visual attention. Support for this interpretation was found in the descriptive narratives. In several instances, the infant basic smiles while studying the pictures, yet both Duchenne and duplay smiles occur when the infant is gazing away from the picture or when she is gazing at and examining the book with other sensory modalities, such as touching the pictures. Basic smiles may occur during book reading in which the communicative function of the smile is enjoyment of the activity, yet the infant can continue to attend and concentrate on the book, given that the cheeks are not raised and therefore not interfering with the infant's vision. Visual attention on the task may play a key

role in the occurrence of basic smiles during parent-infant book reading and object play. Duchenne smiles, which involve cheek raises, may obscure the infant's vision and interfere with the activity; thus, Duchenne smiles may be more likely to occur when the infant is not focusing visual attention on the book. Ekman et al. (1990) found that Duchenne smiles are more often associated with pleasant stimuli and self-reports of pleasure in adults. This may occur because Duchenne smiles partially block the visual input and thereby enhance the personal experience of joyful emotion by relatively reducing input from the environment. It is unclear as to the reason why Duchenne smiles were not significantly related to one specific play type. In light of the complexity of the interaction process gleaned from the qualitative analyses, it is possible that the play categories were too crude to differentiate clear differences for Duchenne smiles in relation to the play types assessed in this study. The finding that smiles with a jaw drop (play and duplay smiles) occur during physical play supports previous research (Blurton-Jones, 1972; van Hooff, 1972). This study assessed both play smiles (lip comer raise and jaw drop) and duplay smiles (lip corner raise, jaw drop, and cheek raise) to determine whether play smiles and duplay smiles occur in different contexts during parent-infant play. The addition of the duplay smile category may have caused the play smile category to be eliminated from the quantitative analyses. Given that this distinction has been ignored in previous research and was obscured in this study, it is imperative for future research to examine the relationship between different contexts and smiles that involve a jaw drop. The descriptive narratives provide evidence that tactile stimulation may be related to the occurrence of play or duplay smiles during physical play. Recall that both play and duplay smiles were assessed in the narratives, and no attempt was made to differentiate them in the interpretation, given the low frequency of play smiles. For example, during the father-infant physical play sequence, the infant dropped her jaw just as the father began tickling her stomach, and it closed again as he lifted his head off her stomach. The unexpected finding that duplay smiles occurred more than expected during object play may be related to tactile stimulation as well. In the mother-infant book reading sequence, the infant's jaw dropped as the book brushed her nose and closed as it was pulled away. This tactile stimulation should not be interpreted as causing the jaw drop. Rather, tactile stimulation may be one factor that helps explain the occurrence of duplay or play smiles during play. Tactile stimulation is incorporated into various types of play through tickling, gently poking, or hugging. An example from the mother-infant object play sequence may help illustrate the point that tactile stimulation inevitably does not cause the jaw to drop: The infant lowers her cheeks and lip corners as the mother pulls back a little when the infant looks at a block. The mother quickly buries her face into the infant's stomach, as she had moments before, yet the infant does not smile this time. The mother attempts again to stimulate the infant's stomach although the infant's attention remains focused on the block that she is now holding. The mother then pulls back her body and watches the infant play with the block.

SMILE TYPE AND PLAY TYPE Other components in the system, such as the infant's increased desire to engage with the block, overstimulation from the tactile component, the mother's facial expressions, temporal patterns, and so forth, may have played a key role in the interaction that did not result in a play smile. Given that tactile stimulation does not inevitably cause or elicit duplay or play smiles, regardless of other components, affords credence to the notion that other components in the system must interact to create the smile types. Smiles are constructed by the interaction of all the communicative components rather than elicited or caused by a specific behavior (Fogel, 1993). Another component that could be influencing the jaw drop during physical play is the fact that the child is more likely to be moving his or her whole body around. The jaw could be similar to other body parts that are exhibiting movement during tactile stimulation. In summary, these findings suggest that duplay-play smiles are more likely to occur during play situations involving tactile stimulation. Basic smiles may occur during more visually demanding activities, such as book reading and some types of object play. Duchenne smiles may occur during less visually demanding situations. It is unclear why Duchenne smiles were not significantly related to a specific play type. The play types assessed may have been too broad to illuminate a specific relationship between play type and Duchenne smiles.

Relationship Between Smile Type, Play Type, and Parent Gender Differences in patterns of communicative behavior were revealed for mother-infant and father-infant dyads as hypothesized. This section provides conjectures as to the meaning of these patterns in parent-infant interaction. The main emphasis will focus on differences in the way in which mothers and fathers engage in the same type of play. Duchenne smiles occurred more often during mother-infant object play, whereas basic smiles occurred more often with father-infant dyads. This difference may be due to the manner in which they engage in play with objects. Previous research (Clarke-Stewart, 1978; Yogman, 1982) has found that mothers engage in object play in a more conventional manner, whereas fathers engage in object play in more idiosyncratic ways. It was suggested previously that basic smiles may more likely occur during visually demanding activities, and Duchenne smiles may more likely occur during less visually demanding activities, as a result of the reduction of visual input during Duchenne smiles (due to the cheek raise). Thus, it may be that fathers' activities, which tend to be more idiosyncratic with objects, require infants to focus their gaze on the object, which helps explain why basic smiles occur. In contrast, mothers tend to use objects in conventional ways, involving patterns of actions that require less visual attention, therefore allowing Duchenne smiles to occur. For father-infant dyads, duplay smiles occurred more frequently during physical play compared with mother-infant dyads. This difference may be due to the different manner in which mothers and fathers engage in physical play (Lamb, 1977; MacDonald & Parke, 1984). Overall, fathers tend to engage in more vigorous physical games that involve more sustained physical contact, such as wrestling and tumbling, and mothers more commonly engage in less high-energy activities that in-

931

volve less vigorous physical contact, such as patty-cake (MacDonald & Parke, 1984). The play interactions with more tactile stimulation may account for the greater frequency of duplay smiles with father-infant dyads. The games in physical play that mothers engage in with their infants may involve less tactile stimulation and require more visual attention; thus, more basic smiles occur during mother-infant physical interactions. The different smile types occurring during physical play may be accounted for by these differences in style of physical play. Duplay smiles were more likely to occur during motherinfant book reading, whereas Duchenne smiles were more likely to occur during father-infant book reading. A reason for this difference is that mothers and fathers may engage in book reading differently with their infants. The descriptive narratives suggest that mothers may vary their behaviors during book reading, such as incorporating objects, physical stimulation, and exaggerated vocalizations into the activity, whereas fathers, who tend to read less with their infants (W. Haight, personal communication, May 22, 1996), may be restricted in their actions. Although this is speculative, evidence from the descriptive narratives suggests that mothers seem to use the book during this activity as a source for further social elaboration. Mothers may use exaggerated sounds and integrate objects into the activity to enhance the book reading activity. The finding that mothers and fathers engage in book reading somewhat differently may help explain the finding that no duplay smiles occurred during father-infant book reading. Duplay smiles may occur during some of the enhanced book reading activity involving tactile stimulation. These findings support the idea that patterns of interaction develop differently for mother-infant and father-infant dyads. Walker (1994), using the same data set, found that the mothers were more likely to alternate between actively interacting with their infants and passively interacting (i.e., helping the infant with a toy and then sitting back to observe the infant manipulate the toy alone). In contrast, father-infant play was more continuously active. That is, the fathers would be continuously interacting or engaging in the infant's activities, without extended periods of passive observation. Walker suggested that mothers and fathers have different intentions when interacting with their infants (H. Walker, personal communication, November 20, 1996). The different interactional sequences or the flow of the interactions may influence the timing of the different smiles and the context in which different smiles occur. Different types of smiles may occur in activities in which the dyad engages frequently, consistent with the notion that patterns develop from the dyad's history of interaction (Fogel, 1993).

Conclusions and Future Research Although this study did not provide any definitive answers as to the meaning of different types of smiles, one could infer that previous interpretations, such as communication of intent and reflection of emotions, are too simplistic to capture the complex relationships between dynamic components of the interaction. Nonobvious perceptual, physical, and cognitive factors could play an important role in different situations that create different types of smiles. No one factor, such as communication intent or emotion reflection, takes precedence over other

932

DICKSON, WALKER, AND FOGEL

components in the creation of different types of smiles. It is important to study the relationship between components of the interaction to determine which components are influencing different facial expressions in different contexts. The information from the narratives was very useful in attempting to generate hypotheses about the complex relationship between smile type and play type. Although beyond the scope of this study, additional questions could be assessed with a combination of quantitative and qualitative analyses. Loglinear analyses incorporating lag sequences could illuminate the temporal ordering of the smile types. For example, do the smile clusters consist of reliable temporal patterns of smile types? Are basic smiles the first phase of a Duchenne or duplay smile? Are these patterns of smile types influenced by context and partner effects? It would be fruitful to examine other explanations that have been raised as well, including the issue of different smile types being related to overall level of activation or arousal (i.e., Demos, 1982) or to cognitive factors (discrepancy resolution), with these techniques. For example, are different types of smiles more likely to occur when the infant anticipates the partner's movement compared to when the infant is not cognitively prepared for the next event in the interactional process? Given the observational nature of this study, causal explanations of the associations are unwarranted. However, the next step, derived from the current findings and from the aforementioned associations, involves experimentation that allows for the examination of causal explanations for the constituents of the communication system. For example, the manipulation of tactile stimulation during play sequences could reveal the function of this variable in the formation of duplay smiles, and the manipulation of visualization tasks could reveal their function in the formation of basic and Duchenne smiles during parent-infant play. In addition, future research could test the speculations that were made regarding the different components of mother- and father-infant play that may help explain the context specificity of smile types.

References Adams, C , Labouvie-Vief, G., Hobart, C. J., & Dorosz, M. (1990). Adult age group differences in story recall style. Journal of Gerontology, 45, 17-27. Andrew, R. J. (1972). The information potentially available in mammal displays. In R. A. Hinde (Ed.), Nonverbal communication (pp. 179— 206). Cambridge, England: Cambridge University Press. Ban, P. L., & Lewis, M. (1974). Mothers and fathers, girls and boys: Attachment behavior in the one-year-old. Merrill-Palmer Quarterly, 20, 195-204. Belsky, J., Gilstrap, B., & Rovine, M. (1984). The Pennsylvania infant and family development project, I: Stability and change in motherinfant and father-infant interaction in a family setting at one, three, and nine months. Child Development, 55, 692-705. Blurton-Jones, N. G. (1972). Non-verbal communication in children. In R. A. Hinde (Ed.), Nonverbal communication (pp. 271-296). Cambridge, England: Cambridge University Press. Cheyne, J. A. (1976). Development of forms and functions of smiling in preschoolers. Child Development, 47, 820-823. Clarke-Stewart, A. (1978). And daddy makes three: The father's impact on mother and young child. Child Development, 44, 466-478. Dedo, J. Y. (1991). Smiling during later infancy: Relationships among

facial expressions, contexts, and other communicative behaviors. Unpublished doctoral dissertation, Purdue University, West Lafayette, IN. Demaris, A. (1992). Logit modeling: Practical applications. (Paper series on quantitative applications in the social sciences, No. 07-086). Newbury Park, CA: Sage. Demos, E. (1982). Facial expressions of infants and toddlers: A descriptive analysis. In T. Field & A. Fogel (Eds.), Emotion and early interaction (pp. 127-160). Hillsdale, NJ: Erlbaum. Ekman, P., Davidson, R. J., & Friesen, W. V. (1990). The Duchenne smile: Emotional expression and brain physiology II. Journal of Personality and Social Psychology, 58, 342-353. Ekman, P., & Friesen, W. (1978). The Facial Action Coding System. Palo Alto, CA: Consulting Psychologists Press. Ekman, P., & Friesen, W. (1982). Felt, false and miserable smiles. Journal of Nonverbal Behavior, 6, 238-258. Feldman, S. S., & Ingham, M. E. (1975). Attachment behavior: A validation study in two age groups. Child Development, 46, 319-330. Field, T, Vega-Lahr, N., Goldstein, S., & Scafidi, F. (1987). Interaction behavior of infants and their dual-career parents. Infant Behavior and Development, 10, 371-377. Fogel, A. (1977). Temporal organization in mother-infant face-to-face interaction. In H. R. Schaffer (Ed.), Studies in mother-infant interaction (pp. 119-151). New \fork: Academic Press. Fogel, A. (1985). Coordinative structures in the development of expressive behavior in early infancy. In G. Zivin (Ed.), The development of expressive behavior: Biology-environment interactions (pp. 2 4 9 267). Orlando, FL: Academic Press. Fogel, A. (1993). Developing through relationships. Chicago: University of Chicago Press. Fox, N. A., & Davidson, R. J. (1988). Patterns of brain electrical activity during the expression of discrete emotions in 10-month-old infants. Developmental Psychology, 24, 230-236. Fridlund, A. J. (1991). Sociality of solitary smiles: Effects of an implicit audience. Journal of Personality and Social Psychology, 60, 2 2 9 240. Green, J. (1988). Loglinear analysis of cross-classified ordinal data: Applications in developmental research. Child Development, 59, 1 25. Hopkins, B., Lems, Y. L., Van Wulfften Palthe, T, Hoeksma, J., Kardaun, 0., & Butterworth, G. (1990). Development of head position preference during early infancy: A longitudinal study in the daily life situation. Developmental Psychobiology, 23, 39-53. Jones, S., & Raag, T. (1989). Smile production in older infants: The importance of a social recipient for the facial signal. Child Development, 60, 811-818. Jones, S., Raag, T, & Collins, K. (1990). Smiling in older infants: Form and maternal response. Infant Behavior and Development, 13, 147165. Kaye, K., & Fogel, A. (1980). The temporal structure of face-to-face communication between mothers and infants. Developmental Psychology, 16, 454-464. Kennedy, J. J. (1992). Analyzing qualitative data: Log-linear analysis for behavioral research. New \ork: Praeger. Kienapple, K. (1987). The microanalytic data analysis package. Behavior Research Methods: Instruments and Computers, 19, 335-337. Knoke, D., & Burke, P. J. (1980). Loglinear models. (Paper series on quantitative applications in the social sciences, No. 07-020). London: Sage. Kraut, R. E. (1978). Verbal and nonverbal cues in the perception of lying. Journal of Personality and Social Psychology, 36, 380-391. Lamb, M. E. (1976a). Interactions between two-year-olds and their mothers and fathers. Psychological Reports, 38, 447-450. Lamb, M. E. (1976b). Twelve-month-olds and their parents: Interaction in a laboratory playroom. Developmental Psychology, 12, 237-244.

933

SMILE TYPE AND PLAY TYPE Lamb, M. E. (1977). Father-infant and mother-infant interaction in the first year of life. Child Development, 48, 167-181. MacDonald, K. B., & Parke, R. D. (1984). Bridging the gap: Parentchild play interactions and peer interactive competence. Child Development, 55, 1265-1277. MacDonald, K. B., & Parke, R. D. (1986). Parent-child physical play: The effects of sex and age of children and parents. Sex Roles, 15, 367-378. Messinger, D. (1994). A dynamic systems perspective on the development of infant smiling. Unpublished doctoral dissertation, University of Utah, Salt Lake City. Messinger, D., Fbgel, A., & Dickson, K. L. (1997). A dynamic systems approach to infant facial action. In J. A. Russell & F. M. Dols (Eds.), The psychology offacial expression (pp. 205-226). New "fork: Cambridge University Press. Oster, H. & Rosenstein, D. (in press). Baby FACS: Analyzing facial movement in infants. Palo Alto, CA: Consulting Psychologists Press. Power, T. G., & Parke, R. D. (1982). Play as a context for early learning: Lab and home analysis. In L. M. Laosa & I. E. Sigel (Eds.), Families as learning environments for children (pp. 147-178). New \fork: Plenum. Smith, W. J. (1985). Consistency and change in communication. In G.

Zivin (Ed.), The development of expressive behavior (pp. 51-75). Orlando, FL: Academic Press. Stevenson, M. B., Leavitt, L. A., Thompson, R. H., & Roach, M. A. (1988), A social relations model analysis of parent and child play. Developmental Psychology, 24, 101-108. Tabachnick, B. G., & Fidell, L. S. (1989). Using multivariate statistics (2nd ed.). New York: Harper & Row. Tisak, M. S. (1986). Children's conceptions of parental authority. Child Development, 57, 166-176. Tisak, M. S., & Turiel, E. (1984). Children's conceptions of moral and prudential rules. Child Development, 55, 1030-1039. van Hooff, J. A. R. A. M. (1972). A comparative approach to the phylogeny of laughter and smiling. In R. A. Hinde (Ed.), Non-verbal communication (pp. 209-241). Cambridge, England: Cambridge University Press. Walker, H. (1994). Parent-infant dyads: Similarities and differences in patterns of interaction. Unpublished doctoral dissertation, University of Utah, Salt Lake City. Yogman, M. W. (1982). Observations on the father-infant relationship. In S. Cath, A. Gurwitt, & J. M. Ross (Eds.), Father and child: Developmental and clinical perspectives (pp. 101-122). Boston: Little, Brown.

Appendix Sample Descriptive Narrative: Father-Infant Physical Play Sequence The infant is rapidly crawling around the floor as the father mirrors her actions, alternating between crawling with his body above hers and crawling slightly behind her. She has a Duchenne smile on her face as she changes directions and the father says lightheartedly, "Don't go that way. Don't go that way." The infant staggers with her head low to the ground before her left elbow bends, resulting in her rolling onto her back. As she rolls, the father lowers his head into her stomach saying "I'm gonna get you." Her cheeks raise and her jaw drops into a duplay smile as his face touches her stomach and she begins to laugh. Her vocalizations fluctuate at the same speed in which he is tickling her stomach with his face and fingers in a side to side fashion. The infant's jaw closes into a basic smile while she looks at him, as the father withdraws his physical stimulation. The smile is gone as she looks away and her vocalizations change in character from laughing to a minor distress call. The father withdraws slightly allowing her to roll back onto her stomach and start crawling again.

Received July 22, 1996 Revision received January 13, 1997 Accepted January 13, 1997

Suggest Documents