Why Do I Hear My Mother When I Open My Mouth?
The Reading Room, Vol. 22 · 8 min read · How much of it actually repeats, which parts carry and which do not, and one sentence to write down
Based on the published research of Sanne B. Geeraerts, PhD — Assistant Professor, Utrecht Centre for Child and Adolescent Studies, Utrecht University, and lead author of the 2025 meta-analysis. Faculty profile · Read the meta-analysis
You heard it leave your mouth and it was not your voice. The research on breaking family patterns found something smaller, and considerably more useful, than the thing most of us are afraid of.
I said a sentence to my niece last spring, in a kitchen, about a coat left on a chair. Eleven words. And the cadence was my mother’s, exactly, down to the small pause before the last two.
My niece did not notice anything. She moved the coat. I stood at the sink with my hands in the water for a good while longer than the washing up required.
What I was doing at that sink, if I am honest about it, was auditing thirty years for other examples. I found four before the water went cold. I am very efficient at this particular task and it has never once helped.
The short answer: breaking family patterns starts with knowing how much is actually being handed over, and the honest number is much smaller than the fear. A 2025 meta-analysis pooled 408 effect sizes and found the average continuity in parenting between one generation and the next sat at r = .12.¹ That is a real relationship and a weak one. It also does not spread evenly. Warmth and sharpness carry more than the rest, and they carry hardest under load, which means the thing you are hearing in your own voice is usually a depletion pattern rather than an inheritance.
The skim version
- Across 24 datasets and 51 published articles, the pooled continuity in parenting from one generation to the next was r = .12, which the authors described as modest.¹
- Continuity was stronger for warmth and for sharpness than for how much independence or structure a parent gave.¹
- It was stronger down the maternal line, and stronger when the children being observed were younger.¹
- A separate UK study found a grandmother’s remembered affection predicted specific behaviours a generation later, with standardised coefficients of 0.18 to 0.26. Grandfathers showed no significant association at all.²
- Both studies measure behaviour. Neither measures character, and neither says anything about what you cannot change.
The source, and who else has measured this
Sanne B. Geeraerts, PhD is Assistant Professor at the Utrecht Centre for Child and Adolescent Studies, Utrecht University, and lead author of the 2025 meta-analysis. Her co-authors include Deborah M. Capaldi, PhD of the Oregon Social Learning Center (profile · Wikipedia · Scholar) and Susan Branje, PhD, Professor of Adolescent Development at Utrecht University (faculty profile · ResearchGate). Faculty profile · Read the meta-analysis · ResearchGate · Google Scholar
Who else has measured this: the smaller UK study cited below, led by Victoria Madden and published in the European Journal of Public Health, independently found the maternal-line pattern in a filmed, coded cohort rather than a pooled meta-analysis.² Neither team has reviewed or endorsed this article.
Do family patterns actually repeat?
Some of them, weakly, and not the ones you would guess.
The largest synthesis of this question was published in 2025. It pulled together 408 separate effect sizes from 24 datasets across 51 peer-reviewed articles, with an average sample of around 373 people each, and asked one question: if you know how somebody was parented, how well can you predict how they parent.¹
The answer was r = .12, with a confidence interval running from .09 to .16.¹
Correlations are a horrible way to talk to a person about their own family, so let me put it in the units you actually use. A correlation of .12 means that knowing your mother’s parenting tells you almost nothing about yours. If it were a weather forecast you would stop checking it. If it were a hiring signal you would remove it from the process. It is not zero, and the researchers found it reliably, which is why the sentence is not “none of it repeats.” The sentence is “much less of it repeats than the woman at the sink believes.”
Which parts carry, and which ones do not
This is where the research gets genuinely useful, because it did not find a flat effect across everything a parent does.
Continuity was stronger on two dimensions in particular: acceptance and negativity.¹ In plainer terms, warmth and sharpness. How available somebody is, and how cutting they get. Those travel.
What travelled considerably less well was autonomy support and structure. Whether a parent lets a child decide things. Whether there is a bedtime and it holds. Those look much more like decisions a person makes than like something handed down, which fits what most of us already suspect, since we have all watched a friend run a household nothing like the one she grew up in.
The effect was also stronger down the maternal line, and stronger when the children being watched were younger.¹ That last detail matters for a woman in her fifties reading this about her adult daughter, because it means the window the research is describing is mostly not the one you are standing in now.
Between us
Two studies sit under everything above and I want you to see their edges. The 2025 meta-analysis is the big one, preregistered, with few indications of publication bias, which is the sort of housekeeping that makes me trust a number more than usual.¹ The UK study is small and deep instead: 146 mothers and 146 fathers, with the grandparents’ parenting recalled by questionnaire and the current generation’s filmed and coded at 24 months.²
Being fair about it: the UK sample was 93 to 95 per cent white, 62 per cent college-educated and better off than average, so it is not a picture of everybody. The grandparents’ parenting was remembered rather than observed, and people remember their childhoods through whatever happened afterwards. The coders agreed with each other about 60 per cent of the time, which is honest of them to report and not brilliant. And here is the one nobody says out loud: almost none of this research is about a fifty-three-year-old speaking to an adult niece in a kitchen. It is about small children. I am extending it, and I am telling you that I am.
If the thing you are actually carrying is a rule about how much you are allowed to need, rather than a habit of speech, that is a different audit. The Quiet Audit is a private ten-minute pass through the rules currently running. Start here.
Why it arrives when you are tired
Because the dimensions that carry are exactly the ones that thin out first when you are running on nothing.
Warmth costs something. Not much, on a good Tuesday. Quite a lot at the end of a fortnight where you have been managing your father’s appointments and a reorganisation at work and somebody’s dog. And when warmth is unavailable, what is left in the drawer is the older, faster script, which is the one you heard eleven thousand times before you were fifteen.
That is why the sentence in the kitchen felt like possession. It was not. It was a shortcut taken by a woman with nothing left to spend on constructing a better one.
| Reading it as inheritance | Reading it as a shortcut | |
|---|---|---|
| What you conclude | This is who I turned out to be | This is what I reach for when there is nothing left |
| What you then do | Audit thirty years at the kitchen sink | Look at what emptied the drawer this month |
| What is available to change | Nothing, which is why it feels so heavy | One specific sentence, in one specific situation |
| What the research supports | r = .12, which is not much of a case | Warmth and sharpness carry, and both are load-sensitive |
None of this means the sentences do not land on the person receiving them. They do. It means the repair work is smaller and more concrete than the story you tell yourself at the sink, and the resentment underneath it usually has its own accounting problem.
The one sentence worth writing down
Write the behaviour, not the identity. That is the whole instruction and it is harder than it sounds.
Not “I am turning into my mother.” That is an identity, it has no edges, and there is nothing you can do with it at eight in the morning. Write instead: the thing I do that she did is, and then finish it with something a stranger could film.
Mine was: I comment on the coat before I say hello. Nine words. Filmable. And once it was on paper it stopped being a character verdict and became a sequence, which meant there was a place to put a different first line.
Do it once, on paper, for one situation. Not a list. A list turns into the audit at the sink, and the audit is the thing that costs you the evening. The family versions of these scripts run on their own rails, and the invisible load underneath them is worth seeing whole.
Questions worth having an answer to
Does this mean my mother’s parenting had no effect on me?
No, and that is not what the number says. It says her parenting is a weak predictor of your parenting behaviour. How you felt growing up, and what you learned to expect from people, is a different question that this research did not ask.
Why is the effect stronger on the maternal side?
The meta-analysis found it, and did not settle why.¹ The plausible explanations range from who does most of the observed caregiving to who gets studied more, and no honest reading of the data picks between them.
My children are grown. Is it too late to change anything?
The research is mostly about young children, so it has nothing to say about you and your adult daughter. What it does suggest is that the dimension most likely to be carrying is warmth, and warmth is not age-restricted.
What if the pattern is my father’s, not my mother’s?
Then the UK study is quietly interesting, because grandfathers’ parenting showed no significant association with the next generation’s behaviour at all.² One small study is not a verdict. It is a reason to hold the paternal version of this fear more loosely than the maternal one.
The one move
Finish this sentence on paper, once, for one situation: the thing I do that she did is. End it with something a stranger with a camera could have recorded. Then put the pen down and do not make a list.
If what surfaces is less a habit than a rule about what you are allowed to need, the Quiet Audit is a private pass through the rules currently running your week. Start here.
Much less of it repeats than the woman at the sink believes.
Informational, not clinical advice. A pattern that feels unmanageable, or a low mood most of the day for weeks, belongs with a therapist or your clinician. Blue Leaf Journal is not affiliated with, and has no commercial relationship with, Dr. Geeraerts, Dr. Capaldi, Dr. Branje or their institutions; none has reviewed, approved or endorsed this article. Checked against the primary sources linked above on 6 September 2026. Corrections: norawhitfield@blueleafjournal.com.
References
1. Intergenerational stability in parenting across two generations: A multilevel meta-analytic review. Psychological Bulletin (2025). 408 effect sizes, 24 datasets, 51 articles. Read the record
2. Madden, V., et al. (2015). Intergenerational transmission of parenting: findings from a UK longitudinal study. European Journal of Public Health, 25(6), 1030-1035. Read the study
Jo Brennan writes the Load desk at Blue Leaf Journal: the admin, the noticing, and the jobs that are technically nobody’s and therefore hers. She wrote her own sentence down in March. She has since said hello first, before mentioning the coat, on four separate occasions, and is choosing to describe that as a trend.
Written by Jo Brennan for Blue Leaf Journal. Updated: 27 August 2026.
