The Forgotten Sample in 'Fight-or-Flight'
When Harvard physiologist Walter Cannon proposed 'fight-or-flight' in The Wisdom of the Body (1932), it quickly became the standard model of human stress response. Sympathetic activation, adrenaline, cortisol, increased heart rate — this cascade became page one of medical textbooks.
In 2000, UCLA social neuroscientist Shelley E. Taylor and colleagues published a paper in Psychological Review asking carefully: 'Whom was this model built on?' Their review found that the overwhelming majority of animal studies Cannon cited used male rodents, and pre-1995 human stress studies had samples that were only ~17% female (Taylor et al. 2000, Psychol Rev, 107, 411–429).
When samples are skewed, conclusions describe a representative subset, not humans in general. Taylor's claim was not 'fight-or-flight is wrong' but 'it shows only part of the stress response.'
Tend-and-Befriend: A Second Pattern
Taylor's evolutionary logic: for pregnant or nursing females, 'fighting' or 'fleeing' would mean abandoning offspring and thus may not be adaptive. So a complementary response of tending (protecting offspring) and befriending (forming coalitions with other females) may have co-evolved.
The candidate neuroendocrine mechanism was oxytocin, released during birth, lactation, and attachment — but also under stress, where together with estrogen it might promote affiliative behavior. Androgens (testosterone) might dampen this effect. The model reached general audiences via The Tending Instinct (2002, Holt), and Geary & Flinn (2002, Psychol Rev) opened the evolutionary-psychology debate.
Side-by-Side: Cannon 1932 vs Taylor 2000
| Dimension | Fight-or-Flight (Cannon 1932) | Tend-and-Befriend (Taylor 2000) |
|---|---|---|
| Trigger | Acute threat, predator, competition | Acute threat + protective obligation |
| Neuroendocrine | Sympathetic, norepinephrine, cortisol | Some parasympathetic + oxytocin, estrogen modulation |
| Behavior | Attack, avoid, flee, freeze | Tend offspring/vulnerable, form coalitions, share info |
| Posited sex difference | Not emphasized (effectively male model) | More common in females (★ later challenged) |
| Evolutionary logic | Individual survival | Offspring survival + group support |
| Clinical implication | Relaxation, calm, block flight | Use social support, caring behaviors |
Evidence Partly Supports, Partly Challenges
Taylor 2006 (Curr Dir Psychol Sci) reviewed self-report data that women seek social support under stress more than men. The decisive counter came from von Dawans, Fischbacher, Kirschbaum, Fehr & Heinrichs (2012, Psychological Science).
They exposed 67 men to the Trier Social Stress Test and then measured trust, cooperation, and punishment behavior. Men too showed significantly increased trust and cooperation after acute stress, with reduced punishment (aggression). The simple 'fight-or-flight' model was inaccurate even for men, and prosocial responses appeared to be a general human option.
Cardoso, Kingdon & Ellenbogen (2014, Psychoneuroendocrinology) meta-analysed oxytocin and stress, concluding that oxytocin's prosocial effects depend heavily on context, relationship, and prior trust. Oxytocin is not the simple 'love hormone' label — it can simultaneously increase in-group trust and out-group wariness, as Insel (2010, Neuron) and others warned.
Critique: The Slide Into Gender Essentialism
The biggest risk of this model is to dress essentialist stereotypes — 'women care, men attack' — in scientific clothing. Eagly and Wood's 'gender similarities hypothesis' line of work has shown that most psychological variables show small or near-zero sex differences.
Evolutionary 'just-so story' critique also applies: 'because it was adaptive for ancestral women, today's women care more' is hard to test directly. Social role theory suggests the same behavioral differences may stem from current social structure and role expectations, not evolution.
Current consensus, roughly:
- Prosocial/caring/coalitional responses are a real mode of stress response (strong evidence).
- The claim that this mode is 'stronger in women' as generalization is weak (mixed evidence).
- Oxytocin is a key candidate but its effects are context-dependent (moderate evidence).
- Culture, role, and relationship matter as much as or more than sex (strong evidence).
Culture and Korean Context
In collectivist cultures, affiliative responses may be stronger regardless of sex. Yoon (2008) and other East Asian studies report that social support seeking is a large share of stress coping.
Korean women's stress research shows a more complex picture. Lee Eun-hee (2008) reports social support as a protective factor in Korean women's work-family stress; Jeong Hye-won (2015) emphasizes family/peer support networks in Korean maternal parenting stress; Kang Hye-jeong (2020) finds peer affiliation networks buffer burnout in Korean working women.
But there is a shadow. In patriarchal society, 'care' may be not a voluntary choice but an enforced role. Caring because the norm 'women must care' is different from one's oxytocin system autonomously selecting that response. Korean discourses around 'K-eldest daughter,' 'sole-burden child-rearing,' and 'in-law emotional labor' show care becoming a depleting, not restorative, resource.
Applying tend-and-befriend to Korea requires a key caveat: care has protective effects only when it is autonomous and reciprocal. Unilateral, enforced caregiving is a risk factor for chronic stress and depression (consistent with Korean data showing female depression prevalence ~1.7x male).
Practice Without the Gender Myth
Using only the strongest evidence:
- Reframe seeking social support as adaptive coping, not weakness. Equally for all.
- Maintain reciprocal relationships. Only-receiving or only-giving relationships weaken the protective oxytocin circuitry (Cardoso 2014 context-dependence).
- Distribute the care burden. Check sex-skewed caregiving in family or organization. Enforced care amplifies, not buffers, stress.
- Normalize affiliation for men. Per von Dawans 2012, men too increase trust and cooperation after stress. The 'men solve alone' myth has weak neuroscience backing.
- Beware oxytocin-as-panacea marketing. Intranasal sprays and supplements lean on 'love hormone' branding with limited clinical effect.
Conclusion: Augment the 100-Year Model, Cautiously
Shelley Taylor's contribution is to have shaken the assumption of a single standard stress response. We can fight or flee — but we can also approach, tend, and befriend. This option is open to the entire species and is not sex-restricted.
At the same time, we know how the simplified message 'women evolved to care' has been used socially. Good science doesn't just convey data accurately; it watches how that data is converted into stories in society.
Next time you are stressed, calling someone — alongside fighting or fleeing — is a response neurobiology endorses. As long as it is your choice, not a gendered obligation.
