Research figure: John Bowlby · Written by Artificial Sapiens
John Bowlby's Attachment and Loss, Volume II: Separation is his 1973 book Separation: Anxiety and Anger, the second part of a three-volume theoretical investigation of attachment, separation, and loss. Its central argument is that anxiety about losing access to an attachment figure, distress during separation, and anger surrounding threatened relationships become intelligible when attachment is understood as a protective behavioral system. Bowlby related ordinary childhood responses to separation to expectations about caregiver availability, evolved responses to danger, and possible developmental pathways toward persistent anxiety. This was a major theoretical interpretation of observational and clinical material, not a contemporary diagnostic system or proof that every childhood separation causes psychopathology. The original volume is documented in the Wellcome Collection Bowlby archive and in the 1973 Basic Books bibliographic record.
Volume II changes the question posed by the first volume. After analyzing how attachment develops and performs a protective function, Bowlby asks what happens when expected access to a trusted person becomes uncertain, is threatened, or is interrupted. He considers why being alone can intensify fear, why repeated threats and inconsistent responsiveness might influence later expectations, and why protest and anger can appear inside a relationship that the child wants to preserve. A historical synthesis in Duschinsky's Cornerstones of Attachment Research shows how these propositions emerged from Bowlby's distinctive combination of separation observations, ethology, clinical psychoanalysis, and a developing model of information processing.
Bibliographic Identity and Place Within the Trilogy
The work was published in 1973 under the title Separation: Anxiety and Anger, as Volume II of Attachment and Loss. Basic Books issued the New York edition; Hogarth Press published a London edition. Physical extent and pagination vary between printings. Catalogs identify the Basic Books edition as xviii plus 456 pages, and its bibliography and index occupy the concluding pages. The title and publisher are verified by Open Library, the 1973 library catalog record, and the Wellcome archive description. Because later reprints contain different pagination, page-specific claims should always identify the edition used.
Volume I, Attachment (1969), develops attachment behavior as an organized, evolutionarily intelligible system. Volume II applies that framework to anxiety, separation, fear, anger, defensive expectations, and individual differences. Volume III, Loss: Sadness and Depression (1980), proceeds to bereavement, mourning, depression, and related defensive processes. This division of intellectual labor matters. The 1973 book is the primary subject here; detailed analysis of Bowlby's whole trilogy or of the 1980 mourning volume belongs to those books' separate bibliographic intents. The volume sequence is also summarized in the Basic Books catalog description.
The 1973 work is best read as a sustained research argument. It assembles laboratory observations, human developmental observations, clinical case interpretation, and findings from other mammals to explain processes that were often grouped under labels such as dependency, fearfulness, or separation anxiety. Its ambition was explanatory rather than taxonomic: Bowlby wanted to describe what behavior accomplishes, under which circumstances it becomes activated, how earlier experience may shape expectations, and why a response that is adaptive in one context may become rigid or costly in another.
The Central Problem: Availability, Security, and Threat
At the heart of Separation is the question of availability. Bowlby's attachment theory holds that proximity to a familiar, responsive caregiver often supplies protection under conditions of danger, unfamiliarity, pain, illness, or fatigue. The ability to approach that caregiver, and the expectation that the caregiver will respond, regulate whether the child can explore or is instead preoccupied with restoring contact. Separation anxiety therefore concerns anticipated or experienced loss of access to someone on whom safety depends. This account follows the theoretical line from Bowlby's early paper on the child's tie to the mother into the 1973 volume.
This distinction makes several observations understandable. A child may tolerate distance while seeing that a caregiver remains accessible, yet become distressed when the same person disappears unpredictably. The physical distance is only one part of the event: the child's developmental capacities, understanding of the situation, history of reliable return, and availability of another familiar caregiver also matter. Under threat, crying, calling, following, clinging, vigilance, or refusal to leave can all function as attempts to maintain or regain security. Bowlby's explanatory claim concerns their organization around a relationship rather than the mere outward similarity of these behaviors.
The concept of availability contains an important cognitive element. A child can react to the caregiver's anticipated response even before separation has occurred. It also contains an interpersonal element: actual patterns of protection, comforting, and reunion offer information from which expectations can develop. Bowlby did not have the modern tools to quantify every component of this account. His volume nevertheless established the theoretical importance of the difference between an attachment figure being physically absent and a child believing the figure to be inaccessible, unreliable, or about to disappear.
Historical Setting: From Psychoanalysis to Direct Observation
Bowlby was trained in medicine, psychiatry, and psychoanalysis, yet he disputed explanations that located much childhood distress primarily in inner fantasy or frustrated feeding drives. Work with separated and hospitalized children brought the actual social environment into theoretical focus. At the Tavistock Clinic, Bowlby developed a separation research program with James Robertson; Mary Ainsworth was subsequently involved in the broader research environment. Duschinsky traces this institutional and intellectual history using archival papers, including unpublished material that complicates simple retrospective accounts of Bowlby's theoretical development (Duschinsky, 2020).
One unusually influential historical observation arose from a young child's stay in hospital. In the 1952 paper A Two-Year-Old Goes to Hospital, Bowlby, Robertson, and Dina Rosenbluth discussed a filmed hospitalization in which a child was separated from familiar care in an unfamiliar institutional setting. The article identifies Robertson as the filmmaker and describes a detailed waking-day observational record jointly made by Robertson and Rosenbluth. The authors' observations gave the clinical argument a vivid empirical object: real behavior unfolding before, during, and after disrupted contact. The original journal article is the appropriate source for these historical details.
These investigations emerged within a period when hospitals often restricted visiting and separated children from parents during treatment. Their historical and policy significance lies partly in the ethical visibility they gave to children's emotional experience. They were, however, concentrated observations in particular institutional conditions. Differences in illness, pain, the unfamiliar environment, substitutes for familiar caregivers, and family history were not experimentally isolated. The film and case material support the claim that severe separation can be distressing; they cannot furnish universal effect estimates for ordinary, supported separations.
The Intellectual Architecture of the 1973 Book
The book proceeds through three major movements. In the first, usually identified as Security, Anxiety, and Distress, Bowlby examines observed reactions to separation and considers their significance for personality and psychopathology. In the second, An Ethological Approach to Human Fear, he explores the circumstances that elicit fear, including cues to danger and the protective role of accessible attachment figures. In the third, Individual Differences in Susceptibility to Fear: Anxious Attachment, he discusses the developmental formation of expectations, anxious attachment, anger, childhood phobias, self-reliance, and divergent personality pathways. This broad architecture is confirmed by the later historical synthesis of Kobak, Zajac, and Madsen in the Handbook of Attachment and by Duschinsky's chapter on Bowlby.
The organization is conceptually revealing. The subtitle mentions anxiety and anger, but much of the work investigates anxiety and fear, including the role of natural danger cues. Anger receives explicit analysis, notably in the discussion of anger, anxiety, and attachment, yet it was not developed into as elaborate a behavioral-system account as attachment itself. Historical scholarship identifies this unevenness as a real limitation in Bowlby's treatment of aggression and anger (Duschinsky, 2020). The volume should therefore be interpreted as an expansive account of separation and fear with an important but comparatively concentrated treatment of anger.
Separation Responses: Protest, Despair, and Detachment
The historical separation research associated with Bowlby and Robertson distinguished patterns commonly described as protest, despair, and detachment. Protest refers to active efforts to restore contact, including crying, calling, searching, and apparent anger. Despair refers to diminishing active protest alongside withdrawal, sadness, or reduced activity as separation continues. Detachment refers to an apparent reduction in overt attachment signaling, sometimes accompanied by altered responses at reunion. These terms describe patterns noticed in particular prolonged separations, especially those involving unfamiliar institutional care; they do not establish three inevitable chronological stages for every child.
An important interpretive hazard arises when quiet behavior is automatically taken as evidence of recovery. Bowlby and his colleagues argued that reduced protest could coexist with sustained disturbance and changed expectations. Contemporary interpretation requires observing the child's behavior across contexts, the quality of substitute care, the timing and manner of reunions, and the child's developmental level. A single snapshot or a parent-reported absence of crying cannot independently establish psychological security, harm, or detachment. The original hospital case material is available in Bowlby, Robertson, and Rosenbluth (1952).
The significance of this work is historical and conceptual. It encouraged clinicians to attend closely to the child's visible relationships and to the environment surrounding medical care. Its most vivid examples came from conditions that differ greatly from everyday childcare, school transitions, brief separations from parents, or supported time with another familiar caregiver. Those contexts must be studied on their own terms. Using the protest-despair-detachment sequence to predict every child's adjustment would exceed what the underlying observations established.
Fear and Anxiety as Organized Responses to Danger
Bowlby's distinctive move was to relate separation anxiety to a more general protective response to danger. He argued that an infant or young child can have biologically prepared sensitivities to certain conditions: sudden unfamiliar noises, darkness, isolation, unfamiliar environments, and especially being cut off from a familiar protective person. Learning and experience then shape when these cues become alarming and how intensely the child responds. The resulting account combines evolutionary predispositions with individual history rather than making either factor an all-purpose explanation. Duschinsky's detailed archival reading discusses this account of fear, innate cue sensitivity, and the possibility that cues interact (Duschinsky, 2020).
The difference between objective danger and perceived danger was essential. A child may react urgently to separation in an objectively safe setting, because the protective cue of an available caregiver has disappeared. Conversely, familiar signs of caregiver availability may reduce alarm even while the physical environment is unfamiliar. In Bowlby's account, this is understandable: evolutionary protective systems need to operate under incomplete information and on warning signs, rather than waiting for danger to become certain. The strength of the argument is its ability to explain adaptive false alarms as well as responses to genuine threats.
That theoretical insight does not determine the physiology of every anxiety disorder. Human fear learning is affected by temperament, cognitive development, culture, exposure history, broader family conditions, and individual neurobiology. Bowlby's ethological account remains a historical foundation for asking how social safety shapes fear, while precise claims about a particular neural pathway, diagnosis, or treatment require their own contemporary evidence.
Natural Clues, Learned Cues, and Compound Threat
The volume examines the difference between cues that may have been recurrent indicators of danger in evolutionary history and cues learned through personal or social experience. In a child who associates an unfamiliar place with prior distress, that setting may intensify alarm. A caregiver's anxious reaction may communicate danger even when the child has not personally experienced the feared event. Bowlby also considered the possibility that combinations of cues—being alone in darkness, for example—can produce stronger fear than either cue in isolation. This is a hypothesis about the organization and triggering of protective behavior; it should not be treated as a universally measured interaction effect in all children.
The role of attachment here is relational regulation. A trusted person's accessibility changes the child's appraisal of danger and the available means of coping. It also helps explain why the same child's fear can fluctuate rapidly when the situation is reframed as safe or the caregiver returns. The account treats reassurance, exploration, and separation protest as linked elements of a regulatory system rather than isolated symptoms.
Anxious Attachment and Expectations of Caregiver Availability
One of the book's most influential claims concerns individual differences in anxious attachment. Bowlby proposed that children build expectations about whether important caregivers will be present, protective, and responsive when needed. Repeated reliable response can support confidence in exploring and seeking help; recurrent uncertainty or threats of abandonment may contribute to heightened vigilance, demands for reassurance, or angry protest. The book is especially important in developing the idea that early interactions are interpreted, remembered, and used to anticipate later relationships.
Bowlby's internal working models provide a language for these expectations. A model of an attachment figure represents whether that person is likely to be available; a model of the self includes expectations concerning one's worthiness of attention and effectiveness in obtaining comfort. These are working, revisable representations—not literal pictures stored unchanged in the mind. Later researchers formalized and tested different aspects of these ideas, and Bretherton and Munholland's overview in the Handbook of Attachment places them within subsequent cognitive and developmental scholarship.
The historical argument does not imply an irreversible early-life destiny. Children can form meaningful attachments to multiple caregivers, their environments change, and experience can update their expectations. Modern attachment classifications describe patterns observed in specific relationships and assessment settings. They are distinct from a clinical diagnosis and should not be inferred from an isolated instance of crying, clinginess, apparent indifference, or anger.
Anxiety, Caregiver Inconsistency, and Threatened Abandonment
Bowlby examined circumstances likely to make caregiver availability feel uncertain, including prior separations, abrupt changes in care, threats of being sent away, and repeated experiences of unavailability. His explanatory proposal was that anxiety can be maintained by uncertainty about access to protection. In this model, intense proximity-seeking may be a rational response from within the child's expectation of a fragile relationship, even when adults label the behavior dependent or excessive.
But the historical evidence was heterogeneous. Case material and clinical interpretation could suggest mechanisms without isolating the causal effects of a single threat or proving that one family practice inevitably produces an anxious attachment pattern. A contemporary study would need prospectively measured exposure, comparison groups, attention to socioeconomic conditions and temperament, and repeated observation of the relationship. Bowlby's conceptual hypothesis remains influential; the strength of each specific causal claim requires evaluation against later designs.
Anger, Protest, and the Preservation of a Bond
Why does anger appear in a relationship organized around safety and closeness? Bowlby's answer was that anger can accompany frustrated proximity seeking. If an important figure disappears, threatens to leave, or repeatedly fails to respond, anger may express protest and pressure toward restored access. In this reading, anger is sometimes directed toward a valued person precisely because the relationship matters. The emotions of affection, alarm, resentment, and relief can coexist when a threatened bond is restored.
This observation moves beyond a simplistic picture in which attachment means warmth alone. A child can seek a caregiver and resist that caregiver at reunion; an adult can desire closeness while expressing resentment about perceived abandonment. Bowlby's analysis helped make such ambivalent reactions comprehensible within one motivational framework. It also raised difficult questions about the functions and developmental consequences of anger that later attachment scholarship addressed more explicitly.
At the same time, anger has many possible functions and antecedents beyond attachment. Frustration, unfair treatment, temperament, sensory overload, coercive interactions, and broader social learning may produce anger without a threatened attachment bond. Duschinsky's historiographic account emphasizes that Bowlby's handling of aggression was much thinner than his work on the attachment behavioral system (Duschinsky, 2020). This limitation matters particularly when using the book to interpret conflict, aggression, or hostility in modern clinical settings.
From Separation Anxiety to Developmental Pathways
Bowlby attempted to explain how expectations surrounding separation might influence later patterns of behavior. If a person expects important figures to disappear or respond unreliably, vigilance and attempts to control access to others may become persistent. If defensive strategies reduce overt proximity seeking, apparent self-sufficiency may conceal difficult expectations about dependence and comfort. These possibilities reflect a developmental theory of adaptation over time, not a one-to-one prediction from a single event.
His account of pathways to personality development was consequently relational and probabilistic in spirit, although the empirical designs available in 1973 could not fully test the long chains of causation he proposed. Later research has assessed whether attachment patterns correlate with subsequent social competence, behavioral problems, and emotional symptoms. Meta-analytic reviews report associations, with variation by outcome, measurement, and sample characteristics; they do not justify assigning a predetermined psychiatric future to a child on the basis of early attachment classification (Groh et al., 2017; Fearon et al., 2010).
Defensive Adaptation and Apparent Independence
Bowlby's psychoanalytic background remained important in his discussions of defense. A person whose bids for support have repeatedly been rejected may reduce the expression of need, redirect attention, or exclude distressing relationship information from conscious consideration. These descriptions became part of a broader attachment-based conversation about emotion regulation and defensive processing. They were influenced by psychoanalysis while being recast in terms of cognition, experience, and behavioral systems.
It is crucial to distinguish historical conceptual language from measured mechanisms. Defensive exclusion in Bowlby's writing is a theoretical account; it does not, on its own, identify an experimentally established neural process or a clinical category. Nor is apparent independence necessarily defensive. Competent autonomous exploration is expected when a person feels secure; low expressed distress can also reflect individual temperament and context. Researchers require appropriate observations and converging data before interpreting a person's behavior as a defense against threatened attachment.
The Role of Mary Ainsworth and Later Attachment Research
Mary Ainsworth was an independent major contributor who gave attachment science observational and measurement foundations. Her fieldwork and later Baltimore studies emphasized caregiver sensitivity, secure-base behavior, and patterned responses to separations and reunions. The Strange Situation procedure and the early secure, avoidant, and resistant classifications arose from Ainsworth and her colleagues' empirical program, not from Bowlby alone. Bretherton's historical reconstruction shows how their intellectual collaboration combined theoretical and empirical lines of inquiry while retaining distinct contributions (Bretherton, 1992).
Separation episodes in the Strange Situation are standardized observational events designed to reveal relationship organization under mild, controlled stress. They are not simulations of prolonged institutional deprivation, nor do they prove that the assessment causes attachment insecurity. Interpretations depend on the entire sequence, including exploration, reunion, and caregiver behavior. Later classifications, including disorganized attachment, were developed after Bowlby's original volumes and require their own methodological history.
Modern research has developed more precise tools and more diverse samples than those available to Bowlby. A 2024 meta-analysis examining maternal and paternal sensitivity and child attachment included 174 samples and 22,914 dyads; the pooled association between sensitivity and security was r = .25, with evidence concerning both mothers and fathers (Madigan et al., 2024). This finding supports the relevance of responsive care while also indicating that much individual variation remains unexplained by sensitivity alone. It also grounds the move away from identifying one biologically specified mother as the exclusive possible source of security.
What Contemporary Evidence Supports
The enduring empirical importance of Bowlby's work lies in a connected set of propositions: children form selective bonds with familiar caregivers; availability and sensitive response matter for the organization of those relationships; threatening or severely neglectful care can disrupt security; and attachment-related differences are associated, on average, with aspects of later adjustment. These findings come from many methodologies developed after 1973. Evidence has strengthened the overall framework while requiring revision of several early assumptions about measurement, specificity, and long-term prediction.
The 2010 meta-analysis by Fearon and colleagues combined 69 samples involving 5,947 children and estimated a small-to-moderate association between insecure attachment and externalizing problems (d = 0.31). The effect differed across sample and measurement conditions, and different attachment classifications were not interchangeable (Fearon et al., 2010). Such findings support prospective risk research without making early attachment a psychiatric diagnosis or a deterministic explanation of a particular child's behavior.
The 2017 synthesis by Groh and colleagues reviewed meta-analytic evidence linking early attachment to social competence, externalizing, and internalizing outcomes, while specifically calling for improved measurement models, mechanisms, moderators, and multisite replication (Groh et al., 2017). Its conclusions are consistent with a developmental framework in which attachment matters but interacts with many other influences. Stronger claims about a particular type of childhood separation demand studies designed to measure separation circumstances and their consequences directly, rather than treating studies of attachment classification as interchangeable with studies of separation exposure.
Multiple Caregivers and the Limits of Maternal Exclusivity
The 1973 volume used language and examples shaped by mid-twentieth-century ideas about mothers and family life. Historical accounts sometimes suggest an indispensable exclusive maternal relationship. Contemporary science instead examines the qualities and continuity of caregiving relationships, recognizing that children may have important bonds with fathers, grandparents, foster parents, adoptive caregivers, and other consistent adults. The 2024 caregiver-sensitivity meta-analysis is directly relevant because it includes both maternal and paternal relationships and finds significant associations for each (Madigan et al., 2024).
The more scientifically precise question is whether the child has stable, responsive, trusted access to protective care, and how changes in that care occur. Severe deprivation, frightening care, unpredictable abandonment, ordinary changes in a childcare schedule, and brief separations with reliable reunion are different exposures. Grouping them together under a single label of maternal deprivation obscures mechanisms and invites unjustified conclusions about families. Michael Rutter's contemporaneous and subsequent critiques demanded more exact differentiation of such experiences (Rutter, 197290005-0); Rutter, 1979).
Scientific Criticism and Methodological Limits
A major limitation of the original clinical evidence is selection. Severe cases arriving in psychiatric or institutional settings are informative about distress and care environments, but they are not representative samples of all children experiencing separation. Illness, social disadvantage, prior adversity, family instability, and the treatment setting may co-occur. The studies that made disruption visible did not consistently distinguish these variables. This restricts causal interpretation even when observations are careful and behaviorally rich.
A second limitation concerns the explanation of anger. Bowlby's attachment framework gives a plausible reason why frustrated proximity seeking could produce anger, but it does not supply a fully developed independent theory of aggression. Duschinsky's study of the archival record stresses the imbalance between the book's subtitle and its actual attention to different emotions, as well as later researchers' concerns about aggression's uncertain place in the theory (Duschinsky, 2020). Any contemporary account linking attachment insecurity to aggression therefore needs to consult modern observational and longitudinal research rather than relying solely on this volume.
A third limitation is the movement from developmental observations to clinical interpretation. Bowlby discussed phobias, agoraphobia, and apparently excessive dependency in the conceptual language of his time. His arguments generated hypotheses about how early relationship expectations might influence vulnerability to anxiety, but clinical diagnosis requires symptom duration, impairment, differential assessment, and contemporary criteria. A case interpreted through a 1973 psychoanalytic and ethological model cannot retrospectively validate a modern diagnostic mechanism.
A fourth limitation arises from cultural and institutional context. The book's examples came primarily from particular Western clinical settings and social arrangements. Meanings of separation, patterns of shared caregiving, opportunities for reunion, and expectations of autonomy can vary. Generalizable claims require cross-cultural sampling, sound measurement equivalence, and attention to the child's own ecology. Further methodological issues concern whether attachment is classified similarly across populations, how repeated assessments are compared, and how temporary states differ from persistent relationship patterns.
Finally, retrospective and narrative coherence can mislead. A convincing developmental story connecting childhood loss to adult anxiety or avoidance may be clinically meaningful as a hypothesis. It does not establish that the earlier event was a unique cause, or that no alternative pathway could produce the same adult outcome. Contemporary causal claims require measured timelines, competing-explanation tests, adequate samples, and, where ethically possible, intervention or natural-experiment designs.
Separation Anxiety as Experience and as Clinical Disorder
Anxiety during separation is a normal developmental phenomenon in many circumstances. The distress may reflect developmental stage, illness, unfamiliarity, or a recent change in care. Separation anxiety disorder is a separate contemporary clinical category involving excessive or developmentally inappropriate fear or anxiety concerning separation from attachment figures, evaluated according to established diagnostic criteria, persistence, impairment, and differential assessment. The historical 1973 model helps explain why separation can become salient, but it is not a diagnostic manual.
A clinician evaluating severe separation anxiety may need to consider the developmental stage, caregiver relationships, recent losses, adverse experiences, school or social functioning, other anxiety problems, trauma, neurodevelopmental differences, physical symptoms, and family circumstances. Decisions about treatment require evidence-based current guidance. For example, the American Academy of Child and Adolescent Psychiatry lists a 2020 clinical practice guideline for anxiety disorders in children and adolescents, which belongs to modern clinical assessment rather than Bowlby's original 1973 argument.
The book's conceptual vocabulary should also remain distinct from attachment disorders associated with severe insufficient care, attachment classifications in the Strange Situation, and the ordinary distress of a child adjusting to a new setting. Each refers to a different level of scientific description. A child who cries during a school transition has not thereby received an attachment diagnosis; a child who appears calm when separated has not thereby demonstrated secure attachment. Proper interpretation depends on the child's history, pattern of functioning, and observation across contexts.
Scientific and Clinical Legacy
Separation changed the questions researchers and practitioners asked of childhood distress. It suggested that care arrangements be evaluated partly through the child's capacity to maintain secure contact and obtain comfort. In pediatrics and institutional care, the separation-research tradition helped focus attention on visits, familiar caregivers, preparation for procedures, consistent substitute care, and the child's experience of reunion. The book also contributed to the language used in developmental psychopathology, emotion regulation, caregiving research, and attachment-informed psychotherapy.
Those applications require proportionate attribution. The historical importance of Bowlby's framework does not mean that every modern intervention derives directly from the 1973 text or that invoking attachment terminology makes an intervention effective. Attachment-based treatment approaches are now evaluated through their own evidence base. A review of attachment-based treatments for adolescents, for example, describes a secure-cycle framework linking representations and attuned communication, and emphasizes the need to define and evaluate treatment processes (Kobak et al., 2015).
The book's influence is also conceptual: it connects expectations of others, fear regulation, defensive adaptation, and the conduct of relationships in a framework that can be examined empirically. Its strongest legacy is the research program it helped generate. Modern evidence provides refined estimates and revised concepts that are richer than any single interpretation available in 1973. Recognizing the original thesis alongside its later testing permits the work to be read as a foundation of a scientific tradition rather than a timeless clinical rulebook.
How to Read Separation Today
A productive reading begins with three questions. First, what precise event is Bowlby trying to explain: temporary inaccessibility, prolonged enforced separation, loss of a previously reliable caregiver, a threat of abandonment, or a persistent expectation that care may fail? Second, what type of evidence does he offer: direct observation, animal comparison, a clinical vignette, a theoretical derivation, or a later developmental extrapolation? Third, which component has actually been tested by subsequent longitudinal, meta-analytic, or intervention research? These questions protect the intellectual specificity of the book.
The volume should also be read in conversation with its historical setting. Its strongest observations came from children's reactions within environments in which care and visiting arrangements could be deeply restrictive. Its theoretical structure reflects both a departure from and continuing engagement with psychoanalysis. Its scientific content has been extended and criticized by later researchers, including those examining caregiver plurality, attachment measurement, culture, anxiety, defensive processes, and the limits of deterministic developmental narratives. The archival reconstruction in Cornerstones of Attachment Research is especially valuable for separating Bowlby's published rhetoric from the evolution of his working ideas.
Why Is the Second Volume Called Separation: Anxiety and Anger?
The title identifies the emotional problems the book seeks to explain after Volume I establishes attachment itself. Anxiety is central because losing access to an attachment figure can undermine expected safety; anger becomes relevant when access is blocked or a valued relationship is threatened. The text nonetheless gives sustained attention to fear and to individual differences in anxiety, with a shorter explicit discussion of anger. The subtitle is an indication of the theoretical problem rather than a promise of equal treatment of two emotions.
Does Bowlby Claim That All Separations Harm Children?
The volume gives particular weight to severe and distressing experiences, including separations involving unfamiliar institutional care and uncertain access to attachment figures. It does not supply valid evidence that every brief or supported separation causes damage. Duration, developmental level, previous experiences, substitute care, predictability, reunion, and broader conditions are necessary dimensions of evaluation. Rutter's analyses emphasized why distinct forms of deprivation and relationship disruption must be examined separately (Rutter, 197290005-0); Rutter, 1979).
Is the 1973 Book the Source of the Strange Situation?
The book builds on observations of separation and reunion, but the Strange Situation procedure and its early attachment classifications belong to Mary Ainsworth's empirical research program. Bowlby's theoretical contribution and Ainsworth's methodology are connected and historically distinct (Bretherton, 1992). The classifications are observational research constructs, not psychiatric diagnoses.
What Is the Difference Between Volume II and Volume III?
Volume II investigates separation, fear, anxiety, anger, expectations of caregiver availability, and potential developmental pathways. Volume III, Loss: Sadness and Depression, focuses more directly on bereavement, mourning, depressive responses, and related defenses. The distinction is historically and bibliographically important even where the psychological processes overlap. The volume sequence is verified by the trilogy's publisher records.
Conclusion
John Bowlby's Attachment and Loss, Volume II: Separation is a landmark in the psychological study of threatened attachment. It reframes separation anxiety, fear, anger, and patterns of apparent dependence or defensive self-reliance in terms of access to protective relationships and learned expectations about their availability. Its historical evidence is indispensable to understanding the emergence of attachment theory, while its generalizations must be assessed with the methodological standards and empirical findings developed since 1973. The book is most accurately understood as an ambitious original account of mechanisms and developmental possibilities whose enduring contribution lies in the scientific questions it made possible to investigate.
Related Articles
John Bowlby — The scientific biography, primary works, and broader intellectual development of the attachment theorist.
Attachment in Infancy and Early Childhood: Security and Development — Current attachment research, observational classification, responsive caregiving, and developmental outcomes.
John Bowlby's Attachment and Loss Trilogy — An integrated guide to Bowlby's three-volume project.
John Bowlby's Separation Response: Protest, Despair, and Detachment — The protest, despair, and detachment separation model.
References
Bowlby, J. (1958). The nature of the child's tie to his mother. International Journal of Psycho-Analysis, 39, 350–373. PubMed.
Bowlby, J. (1973). Separation: Anxiety and anger. Attachment and Loss, Vol. 2. Basic Books. Open Library bibliographic record.
Bowlby, J., Robertson, J., and Rosenbluth, D. (1952). A two-year-old goes to hospital. The Psychoanalytic Study of the Child, 7, 82–94. https://doi.org/10.1080/00797308.1952.11823154.
Bretherton, I. (1992). The origins of attachment theory: John Bowlby and Mary Ainsworth. Developmental Psychology, 28, 759–775. https://doi.org/10.1037/0012-1649.28.5.759.
Duschinsky, R. (2020). John Bowlby and the Tavistock Separation Research Unit. In Cornerstones of Attachment Research. Oxford University Press. https://doi.org/10.1093/med-psych/9780198842064.003.0001.
Fearon, R. P., Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., Lapsley, A.-M., and Roisman, G. I. (2010). The significance of insecure attachment and disorganization in the development of children's externalizing behavior: A meta-analytic study. Child Development, 81, 435–456. https://doi.org/10.1111/j.1467-8624.2009.01405.x.
Groh, A. M., Fearon, R. M. P., van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., and Roisman, G. I. (2017). Attachment in the early life course: Meta-analytic evidence for its role in socioemotional development. Child Development Perspectives, 11, 70–76. https://doi.org/10.1111/cdep.12213.
Kobak, R., Zajac, K., Herres, J., and Krauthamer Ewing, E. S. (2015). Attachment based treatments for adolescents: The secure cycle as a framework for assessment, treatment and evaluation. Attachment and Human Development. https://doi.org/10.1080/14616734.2015.1006388.
Madigan, S., Deneault, A.-A., Duschinsky, R., Bakermans-Kranenburg, M. J., Schuengel, C., van IJzendoorn, M. H., Ly, A., Fearon, R. M. P., Eirich, R., and Verhage, M. L. (2024). Maternal and paternal sensitivity: Key determinants of child attachment security examined through meta-analysis. Psychological Bulletin, 150, 839–872. https://doi.org/10.1037/bul0000433.
Rutter, M. (1972). Maternal deprivation reconsidered. Journal of Psychosomatic Research, 16, 241–250. https://doi.org/10.1016/0022-3999(72)90005-090005-0).
Rutter, M. (1979). Maternal deprivation, 1972–1978: New findings, new concepts, new approaches. Child Development, 50, 283–305. PubMed.
Wellcome Collection. (n.d.). Attachment and Loss, Vol. 2: Separation: Anxiety and Anger (1973), PP/BOW/K.5. Archival catalog.