Why Newborns Look More Like Their Father: The Evolutionary Theory
"He looks just like his dad!" — it's an almost ritual phrase in the maternity ward. The idea that a newborn looks more like their father comes from a specific evolutionary theory. Here's what's confirmed, what's contested, and what to take away.
Daly and Wilson's evolutionary hypothesis (1982)
The idea comes from two Canadian psychologists, Martin Daly and Margo Wilson. Their 1982 thesis: in human evolution, paternity certainty is never absolute. For a father to invest durably in a child (protect, feed, defend), he needs paternity signals.
According to their hypothesis, a baby who visibly resembles his father triggers stronger paternal investment. Natural selection would therefore have favoured a slight pronounced paternal resemblance at birth.
The hypothesis was elegant. Whether it's true is another matter.
The Christenfeld and Hill study (1995) that started it all
In 1995, two researchers published in Nature a study that seemed to confirm the hypothesis. They showed judges photos of babies (at 1, 10 and 20 years old) to compare with photos of parents. Result: judges correctly identified a 1-year-old's father in 49% of cases, vs. 25% expected by chance.
The study made headlines. It was picked up by the press and cited as proof of the evolutionary theory.
Discover who you really look like
Try the Family Plan — €4.99The replications that pushed back
But in science, a single study is never the end of the debate. Several teams tried to replicate the result with larger samples:
- Bredart & French (1999), on 220 photos: no significant preference for the father.
- Brédart & Devue (2008), French study: no marked paternal resemblance.
- Alvergne et al. (2007), on Senegalese families: balanced resemblance between both parents.
Current consensus: the effect, if it exists, is small and inconsistent. The evolutionary theory remains an attractive hypothesis, but the evidence is thin.
The real phenomenon: a shared perceptual bias
What is well-established, however, is a perception and discourse bias. At birth, the maternal entourage tends to emphasise resemblance to the father.
Why? Several explanations:
- Reassuring the father and reinforcing his bond with the child (social utility, regardless of actual resemblance)
- Social conformity: it's what gets said, so we say it too
- "Neutral" newborn profile: a newborn's face is poorly differentiated, making it "interpretable" in any direction
In other words, the baby doesn't resemble the father more than the mother — it's the discourse around the baby that's biased.
And after a few months?
All researchers agree on one point: after 3-6 months, the "looks like dad" effect disappears completely. As the baby's face defines itself, resemblance distributes between both parents — sometimes tilting toward the mother, or even toward a grandparent.
If you really want to know who your child looks like, wait 18-24 months. Before then, half of what you're told is a social projection, not an observation.
Measuring objectively with AI
To get past impressions, you can use a facial comparison tool. Our Look Like Me LLM Engine compares three photos and gives a numerical resemblance score with each parent. Run it again at 6 months, 1 year, 2 years, 5 years: results can evolve significantly.
The most enlightening experiment: test with a photo of the father and mother taken at the same age as the baby. The father rarely wins — and it's often a real surprise.