The origins of 12-month attachment: A microanalysis of 4-month mother–infant interaction

Abstract

A microanalysis of 4-month mother-infant face-to-face communication revealed a fine-grained specification of communication processes that predicted 12-month insecure attachment outcomes, particularly resistant and disorganized classifications. An urban community sample of 84 dyads were videotaped at 4 months during a face-to-face interaction, and at 12 months during the Ainsworth Strange Situation. Four-month mother and infant communication modalities of attention, affect, touch, and spatial orientation were coded from split-screen videotape on a 1 s time base; mother and infant facial-visual “engagement” variables were constructed.

 We used contingency measures (multi-level time-series modeling) to examine the dyadic temporal process over time, and specific rates of qualitative features of behavior to examine the content of behavior. Self-contingency (auto-correlation) measured the degree of stability/lability within an individual's own rhythms of behavior; interactive contingency (lagged cross-correlation) measured adjustments of the individual's behavior that were correlated with the partner's previous behavior.

 We documented that both self- and interactive contingency, as well as specific qualitative features, of mother and infant behavior were mechanisms of attachment formation by 4 months, distinguishing 12-month insecure, resistant, and disorganized attachment classifications from secure; avoidant were too few to test. All communication modalities made unique contributions. The separate analysis of different communication modalities identified intermodal discrepancies or conflict, both intrapersonal and interpersonal, that characterized insecure dyads.

 Contrary to dominant theories in the literature on face-to-face interaction, measures of maternal contingent coordination with infant yielded the fewest associations with 12-month attachment, whereas mother and infant self-contingency, and infant contingent coordination with mother, yielded comparable numbers of findings. Rather than the more usual hypothesis that more contingency is “better,” we partially supported our hypothesis that 12-month insecurity is associated with both higher and lower 4-month self- and interactive contingency values than secure, as a function of mother vs. infant and communication modality. Thus, in the origins of attachment security, more contingency is not necessarily better.

 A remarkable degree of differentiation was identified in the 4-month patterns of “future” C and D infants, classified as resistant and disorganized, respectively, at 12 months. The central feature of future C dyads was dysregulated tactile and spatial exchanges, generating approach-withdrawal patterns. The intact maternal contingent coordination overall safeguards the future C infant's interactive agency. However, future C infants likely come to expect maternal spatial/tactile impingement, and to expect to “dodge” as mothers “chase.” They managed maternal touch by tuning it out, sacrificing their ability to communicate about maternal touch. They “approached” by vigilantly coordinating their facial-visual engagement with maternal facial-visual engagement, but they “withdrew” by inhibiting their facial-visual engagement coordination with maternal touch. We proposed that future C infants will have difficulty feeling sensed and known during maternal spatial/tactile impingements.

 The central feature of future D dyads is intrapersonal and interpersonal discordance or conflict in the context of intensely distressed infants. Lowered maternal contingent coordination, and failures of maternal affective correspondence, constituted maternal emotional withdrawal from distressed infants, compromising infant interactive agency and emotional coherence. The level of dysregulation in future D dyads was thus of an entirely different order than that of future C dyads. We proposed that the future D infant represents not being sensed and known by the mother, particularly in states of distress. We proposed that the emerging internal working model of future D infants includes confusion about their own basic emotional organization, about their mothers' emotional organization, and about their mothers' response to their distress, setting a trajectory in development which may disturb the fundamental integration of the person.

 The findings have rich implications for clinical intervention, with remarkable specificity for different kinds of mother and infant distress. Heightened and lowered self- and interactive contingency, in different modalities, as well as the specific behavioral qualities identified, provide a more differentiated set of concepts to guide clinical intervention.

Keywords:

• mother–infant
• face-to-face interaction
• microanalysis
• attachment

Acknowledgements

This work was funded by NIMH RO1 MH56130, the Koehler Foundation, the American Psychoanalytic Association, the Edward Aldwell Fund, and the Los Angeles Fund for Infant Research and Psychoanalysis; and in part by grants from the National Science Foundation (SBE 0350201) and the National Institute of Child Health and Human Development (RO1 HD 053776) (Lorraine Bahrick). We thank Elizabeth Carlson who coded the Ainsworth Strange Situation videotapes and made an important contribution to the study. We thank Stephen Ruffins and Susan Spieker for Ainsworth reliability coding. We thank Mary Sue Moore, who taught us the most about what attachment means. We thank Karlen Lyons-Ruth, Jude Cassidy, and Mary Sue Moore for their extensive work on the interpretation of results. We thank Amy Margolis, George Downing, Estelle Shane, Frank Lachmann, Doris Silverman, Anni Bergman, Lin Reicher, Steven Knoblauch, Jean Knox, and the Monday afternoon study group, for their consultations on the manuscript. We thank our students who filmed the mothers and infants: Lisa Marquette, Caroline Flaster, Patricia Goodman, Jill Putterman, Limor Kaufman-Balamuth, Elizabeth Helbraun and Shanee Stepakoff. We thank our students who coded the mother–infant videotaped data: Lisa Marquette, Elizabeth Helbraun, Michaela Hager-Budny, Shanee Stepakoff, Jane Roth, Donna Demetri-Friedman, Sandra Triggs-Kano, Greg Kushnick, Helen Demetriades, Allyson Hentel, Tammy Kaminer and Lauren Ellman. We particularly thank Kari Gray, who created frame-by-frame analyses which generated our “cartoon” illustrations; Alla Chavarga, who finalized all tables and figures and who finalized our “cartoon” frame illustrations; Alina Pavlakos, Julia Reuben, Thomas Flanagan and Robbie Ross, who finalized the manuscript, and Nidhi Parashar, our research coordinator, Kate Lieberman and Brianna Hailey, our lab coordinators. We thank Kara Levin, Adrianne Lange, Joseph McGowan, and Sarah Miller, all involved in tables, figures and editing. We thank our research assistants: Matthew Kirkpatrick, Max Malitzky, Helen Weng, Lauren Cooper. We thank our students and colleagues who contributed: Sara Hahn-Burke, Nancy Freeman, Alan Phelan, Danielle Phelan, Paulette Landesman, Tina Lupi, Michael Ritter, Irena Milentevic, Marina Koulomzin, Jeri Kronen, Jillian Miller, Rhonda Davis, Victoria Garel, Leslie Michael, Jessica Silverman, Glenn Bromley, Robert Gallaghan, Naomi Cohen, Kristen Kelly. We thank our lab assistants: Sarah Temech, Priscilla Caldwell, Jennifer Lyne, Elizabeth White, Yana Kuchirko, Greer Raggio, Hope Igleheart, Jessica Latack, Danny Sims, Seth Laucks, Sara Schilling, Michael Klein, Annee Ackerman, Sam Marcus, Iskra Smiljanic, Jenny Lotterman, Gena Bresgi, Ali Pivar, Christy Meyer, Emily Brodie, Sandy Seal, Claudia Andrei, Nina Finkel, Adrienne Lapidous, Nicholas Seivert, Linda Rindlaub, Abby Herzig, Sonya Sonpal, Imran Khan, Julia Disenko, Michelle Lee, Hwe Sze Lim, Marina Tasopoulous, Emlyn Capili, Jessica Sarnicola, Nancy Richardson, Nicole Selzer, Heather Chan, Eunice Lee and Catherine Man. We thank J.T. Yost, our “cartoon” illustrator. We thank John Burke and William Hohauser of ESPY TV. We thank John Auerbach, our excellent in-house editor, and our two anonymous reviewers, who made the work far better. We thank Howard Steele for his generous collaboration on this project. We note that in our previous publication, Rhythms of Dialogue in Infancy (Jaffe, Beebe, Feldstein, Crown & Jasnow, SRCD Monograph Series, 2001), Mary Sue Moore and Rebecca Warner were regretfully inadvertently omitted from the acknowledgements.

Notes

 1. The current data set is entirely different from that reported in Jaffe et al. (2001).

 2. Global codes of mothering style were coded midrange (“good-enough”), intrusive, and mixed midrange/intrusive. No mothers scored withdrawn. Reliability on 22% of dyads, randomly drawn, coded in three waves to prevent coder drift, produced a mean kappa of .90. For the current analysis, codes were collapsed to midrange (“good-enough”) vs. all others. The 84 subjects with attachment classifications did not differ from the 48 without, in “good-enough” vs. “not good-enough” mothering (χ² = .188, 2 df, p = .665). “Good enough” (vs. not good-enough) mothers showed little disruptive behavior (ANOVA F = 16.399, df = 3127, p < .001) (Lupi, 2009).

 3. Because multilevel models have so many covariates, we used correlations (or chi-square or ANOVA) to test associations between attachment and the ordinalized behavioral scales, thus yielding results more comparable to those of the literature.

 4. A “random effect” is the term used for identifying the differences in a variable (function, or association) among the study subjects. These always include variation in the mean of the dependent variable across the observations, and variation in the variance of the dependent variable across the observations; they usually include variation in the linear change in the dependent variable over time, and in our case it includes between-dyad variation in the auto-regressive effect. A “fixed effect” is the average association across study units (in our case, dyads), just as it would be in an ordinary regression analysis. These average effects will account for some fraction of the random effects, just as ordinary regression analysis predictors account for some fraction of the variance in the dependent variable.

 5. Preliminary analyses estimated the number of seconds over which lagged effects were statistically significant. For each dependent variable, measures of prior self or partner behavior, termed “lagged variables,” were computed as a weighted average of the recent prior seconds, based on these analyses. Typically the prior 3 s sufficed to account for these lagged effects on the subsequent behavior. Across the interpersonal modality-pairings studied, mother (M) was significant at 2–3 lags (2–3 s) for both self- and interactive contingency; evaluation of longer lags yielded non-significant results. Significant infant lags varied: for self-contingency, 5 lags (engagement), 4 (vocal affect), 3 (facial affect, gaze), 2 (touch), and 1 (head orientation); for interactive contingency, 6 lags (M gaze → I gaze), 5 (M facial affect → I facial affect), and 3 (M facial affect → I vocal affect, M engagement → I engagement), and 0 (M spatial orientation → I head orientation). Although some of the above modality-pairings show lags longer than 3 s, the amount of variance accounted for was very small. In the analyses, no more than 3 lags were used, to maintain a consistent sample size; no fewer than 2 lags were used, regardless of significance.

 6. Across all dyads (secure and insecure), in this M spatial orientation – I head orientation pairing, there is a significant bi-directional pattern. The infant interactive contingency carries a positive sign: as mothers move from sitting upright toward forward or loom, infants orient away, from enface toward arch (and vice-versa: as mothers move from loom toward upright, infants orient toward enface). The maternal interactive contingency carries a negative sign: as infants orient away from en face toward arch, mothers sit back, moving from loom toward forward or upright (and vice-versa). But maternal interactive contingency reduces to zero in more educated mothers. In future B mothers, maternal interactive contingency is not significant.

 7. An alternative “slope” measure of intrusive maternal touch was reported in Beebe et al. (2003). By fitting semi-parametric group-based mixture models using SAS Proc TRAJ (Nagin & Tremblay, 2001), we estimated two developmental trajectory groups (high and lower) for maternal touch. We calculated the posterior probabilities of group membership for each individual in the sample, estimating the probability of the individual belonging to each trajectory group. This differentiated mothers whose touch behavior remained relatively stable across the session (86.4%), from those who showed decreasing slope (13.6%), a progressively less affectionate touch. Decreasing slope was associated with nonB (p = .04) and with C (p = .006).

 8. There are two exceptions to this generalization in the infant data: (1) “average” infants across the group did not significantly coordinate engagement, vocal affect and touch with mother touch, and the secure subset is similar, differing only in the presence of significant coordination of infant engagement with mother touch; (2) whereas average infants across the group coordinated engagement with mother engagement, future secure infants did not. Maternal interactive contingency findings are identical with one exception: whereas average mothers across the group coordinated touch with infant engagement, future secure mothers did not.

 9. Infant (I) vocal affect was represented 3 times (pairings 3, 6, 9); I engagement and touch 2 times each (pairings 4, 5; and 7, 9 respectively); I gaze, facial affect and head orientation 1 each. Mother (M) touch had 3 (pairings 5, 6, 7); M facial affect 2 (pairings 2, 3), M gaze and engagement 1 each.

10. For infant (I) self-contingency across all 3 attachment contrasts, engagement was represented in 4 findings, touch in 3 (although both variables were equally represented in the modality-pairings). For I interactive contingency, infant engagement coordination with mother (M) engagement yielded 2 findings, I engagement coordination with M touch 2 findings, and I vocal affect coordination with mother touch one. Of infant (I) interactive contingency findings, I engagement was represented in 4 (of 5), M touch in 3, M engagement in 2. For M self-contingency, spatial orientation yielded 3 findings: gaze 2, and facial affect 1, although facial affect was represented twice in the 9 modality-pairings, and gaze and spatial orientation once. For M interactive contingency (exclusively in the D analysis), M engagement coordination with I engagement, and M touch coordination with I touch, each yielded 1 finding.

11. Pearson correlations show that M engagement is correlated with component items of facial affect (r = .873) and gaze (r = .671). Infant engagement is correlated with component items of gaze (r = .664), facial affect (r = .516), vocal affect (r = .479) and head orientation (r = .454), all p < .001. Thus for mothers, face leads slightly over gaze in the constellation, whereas the reverse is true for infants.

Additional information

Notes on contributors

Beatrice Beebe

Dedication This work is dedicated to Daniel Stern, Lotte Kohler, Ruth and Gilbert Beebe.