The concept of intermittent reinforcement has become one of the more widely cited ideas in discussions of attachment in unhealthy relationships. The underlying behavioral science is real and worth understanding accurately. The popular applications of it frequently overstate what the evidence shows.
What the behavioral research established
B.F. Skinner’s foundational work on reinforcement schedules — conducted primarily in controlled laboratory settings with animals in the 1930s through 1950s — produced a well-replicated finding: the pattern of reward delivery affects how persistent the behavior associated with obtaining that reward becomes.
Specifically, a variable-ratio schedule — reward arriving after an unpredictable number of responses — tends to produce behavior that is more persistent and more resistant to extinction than behavior reinforced consistently. The gambling machine is the standard illustration: the unpredictability of the payout drives continued engagement more than a reliable payout would.
This principle has been studied extensively in human contexts as well. The basic finding — that variable reward schedules produce stronger behavioral persistence — is well-supported within behavioral psychology.
How this gets applied to relationships
The application of this research to romantic relationships is an extrapolation, not a direct finding from relationship studies. It’s a coherent and often-cited model, but it should be understood as that: a model.
The proposed mechanism: in a relationship that cycles between warmth and withdrawal, the periods of warmth may produce a stronger response than consistent warmth would — because they arrive unpredictably, after periods of their absence. The relief and intensity of reconnection after a difficult period could, on this model, create unusually strong pull. The positive periods feel more significant because they’re not guaranteed.
Whether this accurately describes what happens in any specific relationship — and whether this mechanism is primarily what’s driving the attachment — is something no behavioral research has established for real-world intimate relationships at the level of specificity that popular accounts suggest.
Where popular accounts overstate it
“Neurochemical addiction.” Versions of this concept frequently claim that intermittent reinforcement creates a dopamine-driven addiction neurologically identical to drug dependency. This is a significant overstatement. There is relevant neuroscience on reward circuitry, but the specific claim that relationship attachment in unhealthy relationships is neurochemically identical to drug addiction is not established by current evidence. These are related systems, not the same phenomenon.
“The pattern proves the relationship is abusive.” Relationship attachment that’s stronger during or after difficult periods can have multiple explanations. Intermittent reinforcement is one possible contributing factor. Others include sunk cost effects, genuine love alongside genuine harm, practical constraints, fear, and the complex history of long relationships. Treating intermittent reinforcement as the singular explanation for all strong attachment in difficult relationships is an oversimplification.
“Understanding it dissolves the attachment.” Knowing about this pattern does not automatically reduce its pull. The attachment — whatever is sustaining it — is real regardless of whether the mechanism has been named. Understanding may help clarify the decisions being made, but it doesn’t function as a switch.
What the extinction research suggests
Behavioral extinction research shows that a learned response can weaken when reinforcement stops, but translating that laboratory principle into intimate relationships is an inference rather than a treatment conclusion. Creating distance may reduce opportunities for an unpredictable reward pattern to continue, but it cannot be presented as a guaranteed psychological intervention.
Distance or no-contact is also not possible or safe for everyone. Shared custody, employment, financial obligations, stalking risk, legal proceedings, and possible retaliation can all change what is practical. Decisions about contact should account for safety and real-world obligations, not behavioral theory alone.
Sources and Further Reading
- B.F. Skinner. The Behavior of Organisms. Appleton-Century-Crofts, 1938. Original variable-ratio reinforcement research.
- Charles B. Ferster and B.F. Skinner. Schedules of Reinforcement. Appleton-Century-Crofts, 1957. Detailed experimental work on reinforcement schedules and behavioral persistence.
- Note: Claims about dopamine and neurochemical mechanisms are not supported by a source here because the article explicitly identifies them as overstated relative to current evidence. Any source offered for that specific claim should be verified before publication.
This article is for educational purposes and does not provide medical, mental-health or legal advice. Only a qualified professional can diagnose a mental-health condition.