Harry Harlow conducted a series of groundbreaking and highly controversial experiments at the Wisconsin Primate Laboratories during the 1950s and 1960s using infant rhesus monkeys.
These studies were meticulously designed to investigate the true nature of attachment and the affectional bond between an infant and its mother.
Cupboard love or secondary drive theories, posited that infants attach to their caregivers simply because the caregiver acts as a conditioned reinforcer by satisfying basic physiological needs, primarily hunger.
Harlow’s research definitively proved this assumption wrong
Harlow showed that attachment develops due to the mother providing “tactile comfort,” suggesting that infants have an innate (biological) need to touch and cling to something for emotional comfort.
Cloth Mother vs. Wire Mother
Experiment 1
Harlow (1958) separated infant monkeys from their mothers immediately after birth and placed in cages with access to two surrogate mothers, one made of wire and one covered in soft terry toweling cloth.
In one condition, only the wire mother provided food via an attached milk bottle; the cloth mother offered comfort alone. In a second condition, the cloth mother provided the milk.
Both groups of monkeys spent more time with the cloth mother (even if she had no milk). The infant would only go to the wire mother when hungry.
Once fed it would return to the cloth mother for most of the day. If a frightening object was placed in the cage the infant took refuge with the cloth mother (its safe base).
Notably, infants fed by the wire mother showed poorer digestion and more frequent diarrhea, suggesting that the absence of contact comfort created measurable physiological stress.
This pattern supports an evolutionary account of attachment: what infants seek from a caregiver is not primarily the provision of food, but sensitive, comforting physical contact.
Long-term behavioural effects
Monkeys reared exclusively with surrogate mothers, regardless of type, showed marked developmental abnormalities compared to those raised by their biological mothers:
- Heightened timidity and fearfulness in novel situations
- Poor social competence with other monkeys
- Submissiveness and an inability to defend themselves
- Difficulty with sexual behaviour and mating
- Inadequate maternal behaviour in females who later had offspring
Critically, these effects were only permanent in monkeys separated for more than 90 days.
Those removed from surrogates before this threshold and placed in a normal social environment were able to form attachments and largely recover, suggesting a sensitive period for social development.
Experiment 2: Fear Responses and the “Secure Base”
Harlow exposed the infants to alarming stimuli, including a loud, wind-up toy bear beating a drum.
Once reassured by her physical presence, they would cautiously turn to face the threat, using her as a secure base from which to explore it.
This dynamic was demonstrated more systematically through an open-field test, in which infants were placed alone in an unfamiliar room filled with novel objects and toys.
When the cloth mother was present, infants initially rushed to cling to her, but gradually grew bold enough to venture out and explore, returning to her periodically for reassurance before setting off again.
Without her, the picture changed entirely: infants froze, crouched on the floor, cried, and sucked their thumbs compulsively, showing none of the exploratory behaviour seen when she was present.
Crucially, substituting the wire mother provided no comfort whatsoever. The infants behaved as though no mother figure were there at all.
Rhesus Monkeys Reared in Isolation
Harlow (1965) also investigated the long-term developmental effects of maternal deprivation by raising infant monkeys in total social isolation, utilizing a steel chamber he called the “pit of despair”.
He kept some this way for three months, some for six, some for nine and some for the first year of their lives.
He then put them back with other monkeys to see what effect their failure to form attachment had on behavior.
The results showed the monkeys engaged in bizarre behavior, such as clutching their own bodies and rocking compulsively. They were then placed back in the company of other monkeys.
To start with the babies were scared of the other monkeys, and then became very aggressive towards them. They were also unable to communicate or socialize with other monkeys. The other monkeys bullied them. They indulged in self-mutilation, tearing hair out, scratching, and biting their own arms and legs.
The severity of these effects corresponded directly to the duration of isolation.
Monkeys isolated for three months showed the least impairment and showed some capacity for recovery when placed in a supportive social environment.
Those isolated for a full year never recovered. The damage appeared permanent.
The effects persisted into adulthood.
Isolated monkeys were socially inept and unable to mate naturally.
When isolated females were artificially inseminated, they proved wholly incapable of maternal behaviour, neglecting, abusing, or in extreme cases fatally injuring their own offspring.
Harlow concluded that privation, defined as never forming an attachment bond at all, is categorically more damaging than deprivation, and that beyond a certain threshold, its effects cannot be reversed.
Maternal Rejection & Abuse
Harlow’s later experiments examined a question with direct clinical relevance: why do abused children so often remain attached to the very caregivers who harm them?
Behaviourist theory predicted that punishment would weaken a bond, since individuals learn to avoid sources of pain.
The clinical evidence suggested otherwise, and Harlow designed experiments to understand why.
Abusive surrogate experiments
Harlow modified cloth surrogates to incorporate mechanisms of rejection.
Some were fitted to emit blasts of compressed air; others contained blunt spikes that could be extended to physically force infants away.
In one variation, a buzzer was sounded as a warning signal before each air blast.
By behaviourist logic, infants should have learned to use the buzzer as a cue to move away.
Instead, when the buzzer sounded, infants clasped the surrogate more tightly, receiving the blast at full force.
When spikes were extended, infants moved only slightly away, then returned and clung as closely as before once the mechanism retracted.
Why fear intensifies attachment
The infants’ responses reflected a core principle of evolutionary attachment theory: frightening situations do not weaken attachment behaviour, they activate it.
When an infant is alarmed, its biological drive is to seek proximity to its caregiver.
Where the caregiver is also the source of the threat, that drive does not diminish; if anything, the fear amplifies it.
The attachment figure and the source of danger are the same individual, and the infant’s nervous system responds to the fear rather than its cause.
Similar patterns have been observed across species: young animals subjected to aggression from a dominant individual will often seek proximity to that same individual immediately afterward.
The “motherless mothers”
Harlow also examined the parenting behaviour of females who had themselves been raised in total social isolation.
Having had no social or maternal experience, these animals were profoundly impaired as mothers.
They rejected their infants’ attempts to cling, ignored their distress signals, and in a number of cases subjected them to serious physical harm.
Yet even infants who suffered sustained maltreatment from their biological mothers continued to display intense attachment behaviour toward them.
The drive to attach persisted regardless of the treatment received.
Taken together, these findings clarified why early abuse so rarely dissolves the attachment bond.
The bond is not contingent on positive experience; it is triggered by the need for safety, and that need is most acute precisely when the caregiver is at their most dangerous.
Conclusions
Harlow’s experiments converged on a clear central finding: the mother-infant bond develops primarily through contact comfort and physical closeness, not through the provision of food.
Warmth and tactile contact are not incidental to healthy development; they are foundational to it.
Beyond this, the research identified a critical period in early development during which some form of attachment must be formed.
Monkeys that were deprived of an attachment figure during this window sustained lasting emotional and social damage.
Those deprived for shorter periods, within the critical period, showed some capacity for recovery when placed in a normal social environment.
Those deprived beyond it did not recover, regardless of subsequent exposure to mothers or peers.
Harlow also refined his interpretation of what, precisely, caused the damage. The key variable was not maternal deprivation in a narrow sense, but social deprivation more broadly.
When infant monkeys were raised in isolation but given twenty minutes of daily contact with peers in a playroom, they developed into socially and emotionally normal adults.
The presence of other infants, even briefly, was sufficient to support healthy development in the absence of a mother.
This distinction has significant implications:
What the developing infant requires is not a mother specifically, but consistent opportunities for social interaction and physical contact during a sensitive early period.
Where those conditions are met, normal development is possible; where they are not, the consequences appear to be irreversible.
The Impact of Harlow’s Research
Harlow’s research transformed developmental psychology, childcare practice, and research ethics in ways that continue to shape the field today.
Overturning “cupboard love” theories
Before Harlow’s experiments, the dominant view in both psychoanalytic and behaviourist circles held that infants became attached to caregivers simply because they satisfied hunger.
This “secondary drive” or “cupboard love” account treated emotional bonds as a byproduct of feeding.
Harlow’s surrogate mother studies refuted this directly:
Because infant monkeys consistently preferred the cloth mother regardless of which surrogate provided milk, he demonstrated that contact comfort is a primary drive, as fundamental to development as the need for food itself.
Empirical support for attachment theory
Harlow’s findings arrived at a critical moment for John Bowlby, who was facing resistance from the psychoanalytic establishment for his claim that the mother-infant bond was a primary instinct rather than a derivative of feeding.
Harlow’s evidence, alongside Konrad Lorenz’s work on imprinting, gave Bowlby the empirical grounding he needed.
Behaviours such as clinging and seeking a secure base when frightened could now be understood as biologically programmed survival mechanisms rather than learned responses to reinforcement.
Childcare and parenting
The practical implications were substantial.
Throughout the early and mid-twentieth century, parents were routinely advised against picking up or comforting crying infants, on the grounds that excessive physical attention would spoil them.
Harlow’s work dismantled this advice by showing that warmth and physical contact are prerequisites for healthy development, not indulgences.
His findings also challenged the assumption that biological mothers hold a unique claim to effective parenting.
Because nursing proved secondary to contact comfort, it became clear that fathers, adoptive parents, and other caregivers are equally capable of providing the secure attachment an infant needs.
Institutional care reform
Harlow’s isolation studies demonstrated that meeting an infant’s biological needs, providing food, water, and shelter, is insufficient for normal development if it is not accompanied by human contact and affection.
This finding drove significant reforms in hospitals and orphanages, where custodial care had often been provided with minimal physical or emotional interaction.
The research established that deprivation of contact is itself a form of harm.
Understanding child abuse
In later experiments, Harlow designed surrogate mothers that intermittently rejected or physically punished the infants.
Despite this, the infants returned to cling to the abusive surrogate each time.
This finding helped clinicians understand why abused children so frequently remain attached to, and protective of, abusive caregivers.
The drive for attachment is strong enough to persist even in the face of consistent punishment.
Ethics of Harlow’s Study
Harlow’s work has been criticized.
His experiments have been seen as unnecessarily cruel (unethical) and of limited value in attempting to understand the effects of deprivation on human infants.
It was clear that the monkeys in this study suffered from emotional harm from being reared in isolation.
This was evident when the monkeys were placed with a normal monkey (reared by a mother), they sat huddled in a corner in a state of persistent fear and depression.
Harlow’s experiment is sometimes justified as providing a valuable insight into the development of attachment and social behavior.
At the time of the research, there was a dominant belief that attachment was related to physical (i.e., food) rather than emotional care.
It could be argued that the benefits of the research outweigh the costs (the suffering of the animals).
For example, the research influenced the theoretical work of John Bowlby, the most important psychologist in attachment theory.
It could also be seen as vital in convincing people about the importance of emotional care in hospitals, children’s homes, and daycare.
How did Harlow’s research influence modern animal research ethics?
The undeniable cruelty of these experiments forced society and the scientific community to confront fundamental moral questions regarding animal research.
It sparked a fierce debate over whether humans have the right to subject animals to such potentially harmful and highly distressing situations for the sake of scientific discovery.
While Harlow’s findings provided monumental insights into attachment, the extreme suffering of his animal subjects highlighted the critical need to balance the pursuit of knowledge with the moral imperative to minimize animal distress.
Ethical Guidelines
In response to these historical controversies, professional organizations like the American Psychological Association (APA) and the British Psychological Society (BPS) established exacting guidelines to protect animal welfare and prevent future abuses.
Today, any institution conducting animal research must have its proposals rigorously evaluated and approved by an Institutional Animal Care and Use Committee (IACUC) or an equivalent ethics board before any experiment can begin.
These committees, typically comprising scientists, veterinarians, and community members, ensure that researchers strictly adhere to the “3Rs” of animal research ethics:
- Replacement: Researchers must actively consider all available options to replace the use of live animals with alternative methods, such as computer simulations or video records, whenever those alternatives can fulfill the research objectives.
- Reduction: There is a scientific and moral requirement to use the absolute minimum number of animals necessary to obtain statistically valid results.
- Refinement: Experimental procedures must be carefully designed and refined to minimize any pain, distress, or discomfort experienced by the animals.
Under these modern ethical frameworks, researchers must definitively prove that their study has a clear scientific purpose and that the prospective value of the research to humans or other animals entirely justifies any potential harm.
Furthermore, psychologists must operate under the assumption that procedures causing pain in humans will also cause pain in animals, requiring them to use anesthetics or analgesics whenever possible and to humanely euthanize animals if they are observed to be in severe, unnecessary distress.
Because of these sweeping ethical reforms, Harlow’s original maternal deprivation and isolation experiments would never be approved or replicated by any institutional ethics board today.
References
Harlow, H. F., Dodsworth, R. O., & Harlow, M. K. (1965). Total social isolation in monkeys. Proceedings of the National Academy of Sciences of the United States of America, 54 (1), 90.
Harlow, H. F. & Zimmermann, R. R. (1958). The development of affective responsiveness in infant monkeys. Proceedings of the American Philosophical Society, 102,501 -509.
