Containment is a concept in psychoanalytical object-relation theories that refers to certain dynamics of primal, non-verbal communication between a mother figure and an infant. This dynamic process is central to the development of internalization capabilities, leading to the formation of the “self” [24] . When referring to adults, containment ability is the capacity process, incorporate and relate to experiences, feelings and emotional impressions within the framework of one’s self [25] .

Human infants begin their psychological development with no clear, differentiated boundaries between the self and the outside world, between internal and external experiences. Threatening internal experiences may be projected outside, thus enabling the removal of these threatening parts from within the self into the outer world. These elements might be undefined sensations of hunger, fear, frustration and discomfort that might manifest themselves in crying, restlessness and fits of rage. The mother’s ability to endure these “attacks” and meet the need they represent, rather than denying them or being overwhelmed by them, is a work of containment. Thus, the mother figure serves as a container, a metaphorical holding vessel for processing these projected elements. Emotional functions of containment performed by the mother include mirroring and echoing when she identifies the infant’s needs in terms such as “you’re cold” or “you’re hungry”. She may also engage in metaphorical holding that allows the infant to suspend recognition of his/her own helplessness [26] . Identifying the need behind an infant’s crying and responding to it appropriately, for example, by holding, feeding or nursing the child to sleep, is a form of emotional processing performed by the mother. This processing enables the infant to incorporate these transformed and now less threatening elements back into its forming self, thus learning to identify his/her needs, fears, longings and emotions within his or her self [27] . Retrieving once projected elements back as part of the self is the foundation of identity formation and the capacity for self-containment, enabling the self to grow and develop [28] . These processes, initiating in early stages of life, are the basis of the human ability to communicate, form relationships, and develop a sense of self-esteem [29] [30] . The internalization processes and dynamics of containment that are central to the development of the self-serve as a blueprint for the ability to self-contain, the ability to bear, withhold and process the internal and external experiences of adult life. The climate and quality of early containment experiences are central in establishing one’s level of containment needs, meaning the need for others to process meaningful positive and negative life experiences.

The emotional arena in which early containment dynamics takes place was conceptualized as a potential space by Donald Winnicott [26] . He defined this space as a metaphorical matrix in which psychological holding is enabled and exercised, allowing the maturation and growth of the infant’s self, secured in a symbolic safety net. We investigate the possibility that Facebook serves as a potential space in which people can fulfill their need for containment.

Facebook as a platform for examining containment needs and their fulfillment. As a digital social network, Facebook provides an accessible, virtual space for communicating, sharing and responding to others. Previous studies established the desire for belongingness as a motivation for using Facebook, defining it as the need for intimacy, closeness, friendship and sharing [4] [11] . Facebook emphasizes sharing among its members because it is a vital component of social networks. The virtual space in which Facebook operates is a realm not confined by time, space, or geography. It contains elements of reality and fantasy, which enables play and practice in identity, self-expression and self-formation [12] . Sunden [31] noted that metaphorically, “We write ourselves into being” when writing on a virtual social network site. Similarly, Turkle [32] stated that the screen itself could be perceived as a symbolic mirror in which parts of the user’s self are reflected.

These descriptions resonate in many ways with the concept of potential space, apart from the crucial element of unconditional love associated with mother- infant relationship. Sharing one’s status, a photograph, or a comment are all ways of sharing parts of the self on Facebook. Regardless of its content and nature, each response to a “self” element shared on Facebook is a validation of one’s existence, an admission of recognition by others. Responses may serve as mirroring and echoing, which are functions of containment. Facebook’s connectedness, which enables constant communication with others, might suggest that one is never alone. Can Facebook serve as an arena in which people can express their containment needs? Can these needs be fulfilled in this arena? Using the lens of containment to investigate these questions raises both theoretical and methodological challenges. The concept of containment was developed in the framework of psychoanalytic therapy. It has not been defined in an operational manner. To meet that dual challenge, we bring together the concept of containment and adult attachment theory [33] .

Adult attachment theory derives from Bowlby’s attachment theory, in which early childhood experiences with major caregivers shape one’s relationships in adult life [33] [34] . When emotional needs such as comfort, holding and security are fulfilled by a caregiver, usually the mother, a secure attachment base can develop. A secured attachment base allows oneself to feel worthy of love, and to perceive others as trustworthy [35] . An inconsistent fulfillment of the infant’s psychological and physical needs might result in an insecure attachment base that can be characterized as anxious or avoidant. In an anxious attachment model, fear of being abandoned and an anticipated rejection by others color relationships with an anxious tone [36] . Adults characterized with an anxious attachment style demonstrate a pre-occupation with the responses of significant others and a heightened awareness of the signs of possible abandonment by significant others. They invest energy in preserving their self-presentation and enhancing existing relationships, while seeking affirmation by and proximity to significant others [35] [36] . In contrast, a consistent experience of neglect of one’s emotional and other needs might lead to an avoidant attachment style, in which others are perceived as untrustworthy or undependable, and the self as autonomous and self-sufficient [36] . Adults with an avoidant attachment style seem to be emotionally numb, avoiding closeness and suppressing their needs from significant others to avoid disappointment. They are reluctant to reveal themselves to others and try to satisfy their emotional needs on their own.

Adult attachment theory and containment. Mother-infant dynamics in which mother figure is emotionally and physically tuned to the infant was referred in attachment theory as envelope interaction, a metaphor for the way mother and infant are entwined in meeting infant’s needs and in establishing a secure psychological reality for them [29] . Envelope interaction, which is the core of attachment base forming has been equated with the dynamics of containment [37] . This analogy paves the way for us to integrate attachment theory and the concept of containment theoretically.

Attachment styles can be described in terms of containment needs. Adults with a secure attachment style have enjoyed an emotional environment of satisfactory containment. Thus, they experience themselves as valuable, and can develop trust and confidence in themselves and others, relieved of fears of extreme dependency or abandonment. They can balance internal processing with communication with others to process various experiences. However, when containment needs are not fulfilled, they compromise the development of a secure attachment base. Adults with an anxious attachment style had their containment needs met in a partial or inconsistent manner. Thus, their sense of security is fragile and depends on physical and emotional proximity and the mirroring of a significant other. An inconsistent emotional presence of a mother figure has prevented the sound internalization of the projected elements of the self, resulting in a strong need for others to regulate and process the individual’s emotional experiences, as well as for self-affirmation. Adults with an avoidant attachment style have had their containment needs consistently ignored. This does not mean that their containment needs are low. Rather they have learned to satisfy their own needs and mute their expression of these needs. Attachment style in research today is viewed alongside two dimensions―anxious and avoidant [38] . A secured attachment style manifests in a low score in both anxious and avoidant dimensions, where an avoidant attachment style manifests in a high score in avoidant dimension and a low score in anxiety dimension [39] .

While the concept of containment resides in the theoretical thinking of psychoanalytical treatment, original attachment theory and adult attachment theory are well established in the field of research. Since attachment theory is centered on internalized relations with others, the social network is a fertile ground for examination of differences in Facebook use and motivations among users with different attachment styles.

Attachment style and Facebook use. Oldmeadow and colleagues [36] found a positive correlation between an anxious attachment style and Facebook use. In their study, subjects with an anxious attachment style spent more time using Facebook and expressed more concerns about the responses of others to them than those with a different attachment style. These subjects also tended to use Facebook more when they were worried or overcome by negative feelings such as loneliness, stress or boredom. These findings were supported in a later study [39] . Given that fulfilling basic containment needs is a central part of shaping one’s attachment style, these findings may support this study’s general hypotheses that Facebook use is correlated with one’s level of containment needs. Adults presenting an anxious attachment style have not enjoyed a consistent and satisfactory fulfillment of their containment needs. Thus, the process of establishing one’s self as a stable, whole entity rather than a fragmented being has not been realized. For such people, the approval of others is crucial for validating one’s self. The search for such approval is evident in how they use Facebook.

To examine the possible relationship between Facebook use and containment needs, we also include several parameters that have not been considered in previous studies. The speed of connecting to and posting on Facebook after the occurrence of a negative event is a parameter associated with containment needs. We expect that those with strong containment needs will log onto a social network faster than others after the occurrence of a meaningful event, because the presence of others, even symbolically, is crucial for the processing of emotional experiences. Other factors we will consider are the frequency and duration of using Facebook and a categorization of that use as active or passive. Examples of active and passive include whether one writes on Facebook or post pictures as opposed to just reading the posts of others.

The theoretical link between containment needs and attachment style [40] has never been tested, and would be the first area of investigation in this present study. Based on the literature reviewed above, we posit that:

H1a: A positive relationship will be found between the anxious attachment style and the need for containment.

H1b: A negative relationship will be found between the avoidant attachment style and the need for containment.

Given that containment dynamics are linked to the processing of negative and threatening experiences [27] we expect that:

H2: A positive relationship will be found between containment needs and the anxiety dimension of attachment style and the sharing of negative experiences on Facebook.

We also add two new parameters of Facebook use―the speed of connecting and sharing on Facebook after the occurrence of a negative experience―to this hypothesis. Thus, we posit that:

H3: A positive relationship will be found between containment needs, the fulfillment of these needs, the anxiety dimension of attachment style and the speed of logging onto Facebook after the occurrence of a negative experience.

We also expect that:

H4: A positive relation will be found between containment needs, fulfillment of these needs, the anxiety dimension of attachment style and the speed of posting on Facebook after the occurrence of a negative experience (H4).

The next hypotheses deal with the effects of having one’s containment needs met and the experience of the self.

H5: A positive relation will be found between an experience of containment, the anxiety dimension of attachment style and the perception of a positive change in feeling when using Facebook.

H6: A positive relation will be found between an experience of containment, the anxiety dimension of attachment style and the feeling of the “self” being seen (self-visibility) on Facebook.

A cross sectional design survey was administered throughout 2014 to 275 Israeli students from various universities across the country. The participants had an average age of 28 years (SD = 8) and were mostly female (71%). The 85-item questionnaire included a short version of the ECR-i, an attachment questionnaire [38] , and items about the participants’ behavior on and use of Facebook such as reading and writing posts, the nature of the material they post on Facebook, the changes in feeling they experienced after sharing posts on Facebook, and items assessing their containment needs. The participants indicated the degree to which they agreed with the statements on a 7-point Likert scale ranging from 1, signifying “never”, to 7, meaning “always”, we determined the items to include in the final version of the survey after two preliminary studies.

The containment needs questionnaire. The concept of containment refers to the dynamics of communication that are non-verbal and unconscious. However, the footprints of these early development processes are evident in the ability of adults to process internal and external experiences. This study deals with the elements of containment needs and their fulfillment about which people can be introspective and report to others. A central element of containment needs is the craving to communicate and share one’s external and internal experiences with others to process them emotionally. Thus, this study regards the need for containment as a need for emotional sharing. Some of the items in the questionnaire address sharing behaviors and needs. Examples include: “When something good happens to me, I feel the need to share it with others”; “When something bad happens to me, I feel a need to share it with others” and “I tend to share day to day experiences with others”. Among those with a stronger need for containment, the inability to share with others might evoke feelings of stress, anxiety and tension. We assessed that possibility with the statement: “When I cannot share my feeling and experiences with others, I feel stressed or anxious.” The fulfillment of one’s containment needs prompts feelings of relief, of being visible and of being validated. Sample items measuring this possibility include: “After sharing with others, I feel better” and “When sharing with others, I feel I am seen”.

The preliminary version of the containment questionnaire included 21 items and was administered to 50 Israeli university students from Haifa and Tel Aviv, 42 of whom answered the questionnaire completely. This questionnaire achieved an internal reliability of α = 0.88, but a principal component analysis and Varimax rotation revealed six different components. Perhaps these results were due to the relatively small sample size, but it was clear that further work needed to be done to create a more coherent questionnaire.

In a second pilot, we administered a new questionnaire to 60 Israeli students. We eliminated the items in the first questionnaire that had high internal correlations. For example, we found no differences between items related to one’s family and those related to one’s friends, so we combined them. Items regarding non-sharing behavior mirrored those about sharing behavior and were omitted. We added an item regarding the transformative nature of the containment process to measure how the responses of others help the processing of emotion: “When I share a negative experience or feeling I have with others, the experience or feeling seem less negative than before.” The second version of the containment questionnaire included 12 items and achieved an internal reliability of α = 0.89. Component analysis in Varimax rotation with Keizer normalization suggested three components: sharing behavior (α = 0.795), the need for sharing (α = 0.865) and the effect of sharing (α = 0.816). Some items loaded on more than one component that fit the three content worlds suggested in those components. For example, item number 8 (“Sharing helps me feel better about myself”) deals with the effects of sharing and with sharing behavior. A component analysis revealed that the three components reduced to two, explaining 66.08% of the variance―the need to share or the containment need and the effect of sharing, the experience of being contained, and the fulfillment of containment needs. Eight items included in this component loaded between 0.60 and 0.84 and explained 37.99% of the variance, with a reliability of 0.87. The effect of sharing included four items that loaded between 0.78 and 0.85 and explained 27.08% of the variance with a reliability of 0.90. Table 1 illustrates the results. Expressions of the need for containment and the fulfillment of containment needs on Facebook could be found outside the exclusive containment questionnaire. Examples include items such as the content of what is shared (“I tend to share negative experiences and events on Facebook”), the speed with which the participants logon to Facebook after a meaningful negative or positive event, the change in one’s feelings after sharing and the change in one’s sense of self after sharing (“When I share personal experiences on Facebook, I feel less lonely” and “When I share personal experiences on Facebook, I feel understood”). These questions were

*Rotated component matrix model.

included in the general survey.

Behavior and uses of Facebook. Thirty items were devoted to different behaviors on and uses of Facebook. The first 11 items dealt with habits of logging onto Facebook, such as its frequency and duration, and the speed of logging on and posting on Facebook after an internal or real event. Nineteen other items dealt with the active use of Facebook such as writing, regardless of whether about happy or sad events, posting a photograph, and using the “like” function, and the passive use evident in reading the posts and comments of others.

Change in reported feeling after using Facebook. The participants were asked to report if and how their feelings changed for the better or worse during various activities on Facebook such as after posting a personal status, after reading the posts of others or looking at their photographs, and reading comments about their posts. Their responses were recorded on a 7-point Likert scale in which 1 signified “my feeling worsens a great deal”, 7 signified “my feeling improves immensely” and 4 signified “I don’t experience any change in feeling”. In order to check the factorial structure in the items related to changes in feeling, we conducted a principal components factorial analysis that revealed two factors that explained 65.84% of the general variance (eigenvalue > 1): a positive change in feeling (7 items) and a negative change (2 items related to negative responses and no responses at all to posts).

The experience of the self and one’s visibility on Facebook. We included several items to measure how the participants experienced themselves while using Facebook. Examples include: “When I’m on Facebook, I feel less lonely”; “When I’m on Facebook, I feel I’m being seen by someone” and “When I’m on Facebook, I feel I can express myself”.

Attachment style questionnaire―ECR. The final version of the questionnaire included 16 items from the short version of the ECR [38] . We omitted two of the statements from the original short version after two preliminary studies showed that doing so would not compromise the reliability of the factor structure.