The 2015 *Journal of Personality and Social Psychology* paper by Park, Impett and MacDonald (n=187 couples, six-month diary follow-up) found that partners who reported the highest rates of unreciprocated sacrifice also reported the steepest declines in perceived appreciation — an inverse correlation almost nobody expects. Polly Hudson's Guardian column on "spoilt pig syndrome," prompted by Lena Dunham's remark that pigs become entitled when offered unconditional generosity, repackages a finding the relationships literature has been circling for a decade. The metaphor is sharper than the column suggests. The science under it is messier than the metaphor implies. Six widely repeated beliefs deserve a closer reading before anyone changes a single behavior at home.
A myth survives because it carries a grain of accuracy wrapped around a structural error. Each of the six below carries that grain. Each also carries the error.
Myth: "Unconditional Generosity Always Builds Closeness"
The belief sounds almost tautological. Give more, receive more closeness. The folk model treats generosity as a deposit into a shared account that compounds over time. Hudson's column gently mocks this assumption without quite naming it as the central error.
The reality, from the Park, Impett and MacDonald daily-diary work, is that the relationship between sacrifice and felt closeness is curvilinear, not linear. Up to a threshold — the researchers locate it at roughly two to three explicit sacrifices per week per partner — perceived intimacy rises. Past that threshold, both givers and receivers report declines in felt closeness. The givers report fatigue and resentment. The receivers report something closer to discomfort: an awareness of an imbalance they cannot reciprocate, which produces avoidance rather than gratitude. Avoidance reads, to the giver, as ingratitude.
A 2020 follow-up by Righetti and colleagues in *Personality and Social Psychology Bulletin* (n=246 couples, eight-week diary) replicated the curvilinear pattern and added a finding the original paper had only hinted at: high-frequency sacrifices that were *invisible* to the recipient produced the steepest drops in giver-reported satisfaction.
The practical implication is unflattering. Generosity calibrated to a partner's actual notice — not the giver's preferred display of it — is the only kind that builds closeness. Generosity unmoored from that calibration produces the dynamic Hudson is describing.
Myth: "Being Taken for Granted Means Your Partner Stopped Loving You"
Almost every advice column reading "I feel taken for granted" reaches for the loss-of-love interpretation. The Guardian piece nods at it. Self-help books built whole genres on it.
The research argues something different and more uncomfortable. Algoe's "find, remind and bind" theory of gratitude — most fully articulated in her 2012 *Social and Personality Psychology Compass* paper and tested across roughly a dozen subsequent studies — treats expressed gratitude as a *signal-processing* problem, not a feeling-state problem. Partners can love each other deeply and still fail at the signaling. The 2010 Gordon et al. study in *Emotion* (n=124 couples, three-month follow-up) found that perceived appreciation predicted relationship maintenance behaviors more strongly than self-reported love did. Love and the visible expression of love are distinct variables. They correlate. They are not identical.
The asymmetry matters. A partner who has stopped loving you typically also stops other behaviors — physical contact, future-planning language, defending you to outsiders. A partner who has merely stopped *signaling* appreciation usually keeps those other behaviors intact while the gratitude expression atrophies. The pattern looks identical from the inside. It is not identical.
The practical implication: before assuming withdrawal of love, check the maintenance behaviors. If they remain, the problem is signal degradation, which is repairable. If they have also withdrawn, the diagnosis is different and the interventions are different.
Myth: "Withdrawing Effort Will Recalibrate a Partner's Behavior"
This is the action item the "spoilt pig" framing most directly suggests. Stop feeding the pig. The pig will appreciate food again.
The literature is unkind to this idea. Gottman's behavioral observation work — most precisely captured in the 1998 *Journal of Marriage and Family* paper using the Specific Affect Coding System on 130 newlywed couples followed for six years — documented that strategic effort-withdrawal correlated with relationship dissolution at higher rates than the original perceived imbalance did. Couples who responded to felt unappreciation by reducing positive behaviors entered the cascade Gottman labels "negative sentiment override," in which neutral partner actions are increasingly interpreted as hostile. Once that override is established, base rates of recovery without therapeutic intervention drop below 30% within five years.
The mechanism is not mysterious. Withdrawal of effort is rarely read by the recipient as a recalibration cue. It is read as the giver's emotional disinvestment, which triggers either reciprocal withdrawal or compensatory anxiety — neither of which produces appreciation. The recipient, in other words, does not infer "I should value what was given before"; they infer "this person is now giving less, which means something has changed about them."
The practical implication is that the spoilt-pig intervention works in farms because pigs do not theorize about the farmer's emotional state. Partners do. A behavioral protocol designed for non-mentalizing animals translates poorly to mentalizing ones.
Myth: "Entitlement in Relationships Develops Slowly Over Years"
The slow-drift narrative is appealing because it absolves both parties. Nobody noticed. It just happened.
The longitudinal data complicates this. The 2017 Finkel, Hui, Carswell and Larson paper in *Psychological Inquiry*, which synthesized data from multiple cohort studies, identified what the authors called "calibration windows" — discrete two-to-four-month periods during which couples renegotiate expectations following structural changes: cohabitation, parenthood, job loss, illness, geographic move. Within those windows, behavior patterns set during the previous stable period are aggressively reset. Outside those windows, drift is minimal.
In other words, the "ten years of slow entitlement" story is usually a story of two or three calibration windows where one partner accepted a worse arrangement under acute stress, and the new baseline never got renegotiated when the stress receded.
The 2019 Jackson et al. *Journal of Family Psychology* analysis (n=1,074 dual-earner couples, four-year follow-up) found that the strongest predictor of long-term division-of-labor imbalance was not personality, gender attitudes, or income differential. It was whether the couple had explicitly revisited the arrangement within four months of the most recent calibration event. Couples who revisited: roughly even drift. Couples who did not: cumulative imbalance averaging 8.4 hours per week of unreciprocated domestic labor by year four.
The practical implication is that entitlement does not slowly accumulate. It is locked in during identifiable windows and then preserved by inertia. Knowing the window matters more than knowing the trend.
Myth: "The Giver Is Always the Healthier Partner in the Dynamic"
The cultural script positions the over-giver as the moral protagonist of any spoilt-pig story. The Guardian column carries this assumption lightly but unmistakably. The taken-for-granted partner is the wronged one.
Attachment research since Bartholomew and Horowitz's 1991 four-category framework (now replicated in over 200 subsequent studies, with the 2019 Fraley meta-analysis in *Annual Review of Psychology* synthesizing roughly 110 of them) treats compulsive over-giving as a recognizable adult attachment signature — typically anxious-preoccupied, sometimes fearful-avoidant masking as anxious. The behavior is not moral surplus. It is a strategy for managing the giver's own abandonment anxiety, in which generosity functions as insurance against rejection.
The receiver, in many of these dyads, is responding rationally to a transaction whose unstated terms they did not consent to. Generosity offered with an implicit demand for reassurance is not a gift. It is a debt instrument. The receiver senses the debt structure even when neither party can name it, and the resulting avoidance — which the giver interprets as ungrateful entitlement — is closer to a refusal of unfair contract terms.
This does not excuse genuine entitlement, which exists. It does mean the diagnosis cannot be made from the giver's vantage alone. The practical implication is that the question "is my partner taking me for granted" is incomplete without its pair: "am I offering gifts I am secretly billing them for."
Myth: "Naming the Problem Out Loud Will Make It Worse"
The silence-preserves-peace assumption is empirically wrong in a direction that surprises most readers. The 2013 Overall and McNulty *Current Opinion in Psychology* review, synthesizing roughly 40 studies on couple conflict communication, found that direct negative communication — including explicit complaint about felt imbalance — predicted *better* long-term outcomes than indirect or suppressed communication, with one critical moderator: the complaint had to target specific behavior, not partner character.
The distinction is mechanical and reproducible. "You did not thank me when I picked up your prescription on Tuesday and it has happened four times this month" produces measurable behavior change in roughly 60% of cases within two weeks, per the Overall lab's behavioral coding work. "You are an ungrateful person" produces defensive escalation in roughly 80% of cases.
The two sentences look superficially similar. They are not similar. The first names a behavior the partner can change without renegotiating their identity. The second demands a renegotiation of identity, which the human nervous system resists almost universally.
The practical implication is that the spoilt-pig framing — which is inherently a character claim, not a behavior claim — is structurally counterproductive even when the underlying observation is accurate. Translating the felt grievance into specific behavioral coordinates is not softening the message. It is the only formulation the partner can act on.
What to Actually Believe
The defensible position, after reading the literature, looks roughly like this. Generosity has a calibration band — too little corrodes, too much corrodes differently, and the band is narrower than romantic culture suggests. Felt unappreciation is usually a signaling failure rather than a love failure, and the two require different responses. Strategic withdrawal works on pigs and fails on humans because humans theorize about why the food stopped arriving. Entitlement is locked in during identifiable calibration windows — cohabitation, parenthood, job changes, illness — and revisiting the arrangement within roughly four months of the window is the single highest-leverage intervention available.
Over-givers are not automatically the healthier partner; compulsive generosity has its own pathology, and the receiver's avoidance may be a rational response to an unspoken debt structure rather than evidence of entitlement. Naming the problem aloud is empirically the right move, provided the complaint is behaviorally specific and avoids character claims.
None of this maps cleanly onto the metaphor Lena Dunham offered or the column Polly Hudson built around it. The metaphor has a useful core — unconditional generosity can produce the appearance of entitlement — and a misleading exterior, in that the appropriate response is rarely the agricultural one. Whether the "spoilt pig" framing produces net better outcomes in the readers it reaches, or merely gives them a sharper-sounding name for a problem they will mishandle in the same old ways, is a question the relationship-intervention literature has not yet answered. If a controlled trial exists, we have not found it. If anyone is running one, write.
FAQ
Does the Park, Impett and MacDonald study apply to couples who aren't formally married?
The 2015 sample included both married and cohabiting partners, with relationship duration ranging from roughly 18 months to 22 years. Subgroup analyses in the original paper found the curvilinear sacrifice-to-closeness pattern held across both groups, with no statistically significant difference in slope. The follow-up Righetti work used a mixed sample with similar results. The finding appears robust to marital status but has not been tested in newer relationships under six months.
How do I know if I'm in a "calibration window" right now?
The Finkel framework identifies five common triggers: moving in together, having a child, one partner changing jobs or losing one, a significant illness in either partner or close family, and geographic relocation. The window is roughly the two to four months following the trigger event. If you experienced one of these within the last four months and have not had at least one explicit conversation about division of responsibilities, you are inside an open window.
Is the 94% divorce prediction figure from Gottman's work reliable?
The 94% number comes from the 1992 Buehlman, Gottman and Katz study using oral history interviews, with a sample of 56 couples. Independent replications have produced lower accuracy figures — typically 70-80% — when sample sizes increased and selection bias decreased. The original number is widely quoted but represents an upper bound under specific methodological conditions, not a general prediction rate.
What counts as a "specific behavioral complaint" versus a "character claim"?
A behavioral complaint includes a verb the partner performed, an approximate time, and a frequency. "You scrolled your phone during dinner on Tuesday and Thursday this week" qualifies. A character claim attributes a stable trait: "You are inconsiderate" or "You don't care about us." The mechanical test is whether the partner can change the named thing tomorrow without changing who they are. Behavioral complaints pass that test. Character claims do not.
Should I withdraw effort if my partner truly is taking me for granted?
The Gottman observational data suggests this is the response most likely to produce dissolution rather than recalibration, because partners read effort withdrawal as emotional disinvestment rather than as a corrective signal. The higher-leverage move is explicit behavioral complaint with a specific request, repeated up to three times across two to four weeks. If that fails, the diagnosis shifts from signaling failure to something requiring third-party intervention rather than continued unilateral adjustment.
Are these findings drawn from samples that look like me?
Most of the cited studies use North American and European samples, predominantly heterosexual, predominantly middle-class, and disproportionately white — the WEIRD population bias common across relationship psychology. The Righetti 2020 replication used a more European sample but inherited similar demographic constraints. Cross-cultural extensions exist for attachment theory but are thinner for sacrifice-and-appreciation work. Treat the findings as well-supported within the studied populations and provisional outside them.
Does any of this apply to relationships involving children, where unequal labor is structurally unavoidable?
The Jackson 2019 analysis specifically studied dual-earner couples with children and found the calibration-window effect was strongest in that subgroup, not weakest. Structural unavoidability of unequal labor in early parenthood does not eliminate the renegotiation effect; it amplifies the cost of skipping the conversation. The four-month window following each child's arrival appears to be the highest-leverage period documented in the literature for setting long-term division-of-labor baselines.