You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Having built the machinery of classical and operant conditioning, we can now turn it on one of the most important bonds a human being ever forms: the attachment between an infant and its caregiver. The learning approach offers a bold and economical claim — that this bond, for all its emotional depth, is nothing more than a set of learned associations, acquired by exactly the same processes that make a dog salivate to a bell or a rat press a lever. On this view there is nothing mysterious about a baby's love for its mother: she becomes loved because she is repeatedly paired with the satisfaction of the infant's most basic need, food. This is the so-called "cupboard love" theory, and it was set out most systematically by Dollard and Miller (1950). This lesson explains how classical and operant conditioning are combined to explain attachment, why the theory is elegant, and — crucially — why the weight of evidence, above all Harlow's (1958) monkeys and Schaffer and Emerson's (1964) infants, turns out to undermine it. The story is a model of how a plausible learning-theory hypothesis is tested and found wanting, which is exactly the kind of evaluative narrative the Edexcel exam rewards.
This lesson addresses the Edexcel 9PS0 — Paper 1, Topic 4: Learning Theories content on the learning theory of attachment — the explanation of attachment through classical conditioning and operant conditioning ("cupboard love"), associated with Dollard and Miller (1950) — and its evaluation against contradictory evidence (Harlow's monkeys; Schaffer and Emerson). In assessment-objective terms, you should be able to describe how the two conditioning processes are applied to the formation of attachment (AO1), apply the model to examples of infant–caregiver interaction (AO2), and evaluate it, principally through the animal and human evidence that contradicts it and the cognitive and biological factors it omits (AO3).
Connects to…
The learning approach begins from a simple premise: a newborn does not arrive loving anyone. Attachment, like every other behaviour on the behaviourist account, must be learned — and learned through the two mechanisms already established in this topic. The central idea is that the caregiver becomes associated with the reduction of the infant's discomfort, above all the discomfort of hunger. Because the caregiver is present whenever hunger is relieved, the infant comes to seek out, and take pleasure in, the caregiver's presence. The bond is, in the memorable phrase, based on the caregiver being a kind of walking "cupboard" from which good things (food, warmth, comfort) come.
Key Definition: The learning theory of attachment ("cupboard love") — the explanation, associated with Dollard and Miller (1950), that an infant becomes attached to its caregiver because the caregiver is repeatedly associated with the satisfaction of the infant's primary drives (chiefly hunger), so that attachment is learned through classical and operant conditioning rather than being innate.
This claim rests on the wider behaviourist framework of drives. A primary drive is an innate, biological motivational state — hunger, thirst, the need for warmth — that the organism is compelled to reduce. Drive reduction is inherently reinforcing: any behaviour or stimulus that reduces a primary drive is "stamped in". Attachment, the learning theorist argues, is a secondary drive: it is not innate, but is learned by association with the reduction of a primary drive. Because the caregiver is the reliable agent of hunger reduction, the infant's drive to be near the caregiver is acquired on the back of the hunger drive. Dollard and Miller (1950) set this out within their broader attempt to translate psychoanalytic and developmental ideas into the language of learning theory, and it is their names the Edexcel specification attaches to the account.
The first half of the explanation uses classical conditioning to explain how the caregiver comes to produce pleasure in the infant simply by being present. The analysis maps exactly onto Pavlov's terminology, and being able to lay it out precisely is a reliable source of AO1 and AO2 marks.
| Stage | Stimulus | Response |
|---|---|---|
| Before conditioning | Food (UCS) | Pleasure / satisfaction (UCR) |
| Before conditioning | Caregiver (NS) | No special response |
| During conditioning | Caregiver (NS) + Food (UCS), repeatedly paired | Pleasure (UCR) |
| After conditioning | Caregiver alone (CS) | Pleasure / comfort (CR) |
Food is an unconditioned stimulus (UCS) that naturally produces a feeling of pleasure and satisfaction — the unconditioned response (UCR) — without any learning. At first the caregiver is a neutral stimulus (NS), producing no particular emotional response in the infant. But because the caregiver is present every time the infant is fed, the caregiver is repeatedly paired with food. Over many such pairings the caregiver becomes a conditioned stimulus (CS), now able, on their own, to produce the feeling of pleasure and comfort — now a conditioned response (CR) — even when no food is present. In everyday terms, the baby learns to associate its mother with the good feeling of being fed, so that the mere sight, sound or smell of her comes to produce that good feeling directly. That learned association, the theory claims, is attachment.
graph TD
A["NS: Caregiver — no special response"] --> B["NS + UCS: Caregiver paired with Food"]
C["UCS: Food"] --> D["UCR: Pleasure / satisfaction"]
B --> E["Repeated pairings over many feeds"]
E --> F["CS: Caregiver alone"]
F --> G["CR: Pleasure / comfort — the learned bond"]
style G fill:#27ae60,color:#fff
Classical conditioning explains how the caregiver acquires the power to comfort, but it does not by itself explain why the infant actively seeks out and works to maintain proximity to the caregiver. That is the job of operant conditioning, which explains attachment behaviours — crying, clinging, following, smiling — as voluntary operants shaped by their consequences. Two reinforcement processes work together, one on the infant and one on the caregiver.
For the infant, attachment behaviour is negatively reinforced. A hungry, cold or distressed infant is in an unpleasant, drive-heightened state. When the infant cries and the caregiver responds by feeding, warming and soothing them, the unpleasant state is removed. Because an aversive stimulus (discomfort) is taken away, the behaviours that summoned the caregiver — and, more generally, being close to the caregiver — are negatively reinforced and become more frequent. The caregiver thereby also becomes a secondary reinforcer: through repeated association with the primary reinforcement of food and comfort, the caregiver herself acquires reinforcing value, so that simply being near her is rewarding. Proximity-seeking is thus strengthened directly.
For the caregiver, responding is also negatively reinforced. An infant's cry is an aversive stimulus for the adult too. When the caregiver responds and the crying stops, that aversive stimulus is removed, so the caregiver's caregiving behaviour is negatively reinforced and becomes more likely in future. Attachment, on this account, is therefore a two-way, mutually reinforcing system: the infant is reinforced for seeking proximity and the caregiver is reinforced for providing it, and the bond is built and maintained by this reciprocal exchange of reinforcement. This mutual-reinforcement point is a sophisticated AO1 detail that lifts an answer above a one-sided "the baby gets fed" account.
| Party | Behaviour | Consequence | Type of learning |
|---|---|---|---|
| Infant | Crying, clinging, following | Discomfort (hunger, cold) is removed by the caregiver | Negative reinforcement → proximity-seeking increases |
| Infant | Approaching / staying near caregiver | Caregiver = secondary reinforcer (associated with food) | Positive/secondary reinforcement → attachment strengthened |
| Caregiver | Feeding, soothing, holding | Aversive crying stops | Negative reinforcement → caregiving increases |
The operant half of the theory rests on a piece of behaviourist machinery that is worth unpacking carefully, because examiners reward candidates who can explain why attachment is classed as a learned drive rather than simply asserting that it is. Behaviourist learning theory, in the version Dollard and Miller (1950) developed, holds that all motivation begins with primary drives — innate, biologically given states of tension such as hunger, thirst, cold and pain that the organism is compelled to reduce. Anything that reduces such a drive produces a state of satisfaction, and, by the drive-reduction principle, any behaviour or stimulus reliably present at the moment of drive reduction is reinforced. This is the engine of all learning on the account: organisms learn whatever reduces their drives.
Attachment, however, is plainly not a primary drive — a newborn is not born needing its particular mother in the way it is born needing food. The learning theorist therefore classifies attachment as a secondary drive: a motivational state that is not innate but is acquired by association with the satisfaction of a primary drive. Because the caregiver is present at every episode of hunger reduction, the need for the caregiver is learned "on the back of" the hunger drive. The infant does not merely learn to feel pleasant in the caregiver's presence (the classical component); it develops a genuine motivation to seek her out and remain near her, because proximity to her has become bound up with the reduction of its most pressing biological needs. The caregiver thus becomes a generalised secondary reinforcer — a stimulus that, like money for an adult, has acquired reinforcing power through its association with many primary reinforcers (food, warmth, relief of discomfort). This is the theoretical heart of "cupboard love": the caregiver matters because she is the reliable, learned route to drive reduction.
Laying the mechanism out this way also clarifies exactly where the theory is vulnerable. The whole account depends on feeding (drive reduction) being the causal origin of the secondary drive. If it could be shown that infants form strong attachments to figures who do not feed them, or attach more strongly to a non-feeding source of comfort than to the feeder, the secondary-drive argument would be cut off at its root — because the primary drive it is supposed to be built on would not be doing the work. That is precisely the vulnerability the evidence from Harlow and from Schaffer and Emerson exploits, and understanding the drive-reduction logic is what lets you explain why their findings are so damaging rather than merely noting that they "go against" the theory.
It is also worth contrasting the learning-theory drive with the evolutionary account you will meet later. Where the learning theorist sees the drive to attach as learned from feeding, the evolutionary theorist (Bowlby) sees it as an innate drive selected because infants who stayed close to a protective adult survived to reproduce. The two accounts make different predictions about a crucial case — an infant who is comforted but not fed by a particular figure — and it is exactly this case that Harlow engineered. Holding the two explanations side by side, and seeing that they diverge on a testable prediction, is the kind of synoptic framing that lifts an evaluation from competent to top-band.
An AO2 stem will typically describe an episode of infant–caregiver interaction and ask you to analyse it in learning-theory terms. Consider a six-month-old who begins to cry, is picked up and fed by her father, and settles; over the following weeks she smiles and reaches out whenever he enters the room.
Notice the skill the exam rewards: taking the described behaviours apart into UCS/UCR/NS/CS/CR and reinforcement contingencies, and naming who is reinforced by what. A common AO2 error is to force everything into classical conditioning; the strongest answers show that the learned association (classical) and the proximity-seeking (operant) are two different mechanisms doing two different jobs. Equally, a scenario can be used to challenge the theory: if the stem notes that the infant actually attaches most strongly to a grandparent who plays with her but never feeds her, you can point out that the learning account predicts attachment to the feeder and therefore struggles with the case — a neat bridge from AO2 into AO3.
The theory's greatest strength is that it has genuine explanatory face-validity: feeding really is associated with attachment, and association really does shape behaviour. It is undeniable that infants are fed by their caregivers many times a day and that the two are closely associated, and the underlying conditioning principles are among the best-evidenced in psychology. This matters because it means the theory is not baseless: some element of learning by association almost certainly does contribute to the developing bond, and classical and operant conditioning are demonstrable, testable mechanisms rather than speculation. The implication, however, is that showing feeding is associated with attachment is not the same as showing feeding causes it — a distinction the contradictory evidence below exploits, and one a top-band answer draws explicitly.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.