"I can hear that you're angry. When you're ready to talk, I'm here."
Do this: Remove the reaction. Stay in the room if that's where you already are, but reduce your engagement with the screaming itself to near zero. Keep your face neutral. Keep your body calm. If you need to leave the room to manage your own response, leave. Come back when the screaming has reduced, not before.
Skip this: Arguing with the screaming. Telling them to stop, which is engagement. Offering alternatives or substitutes mid-scream, which teaches them that screaming unlocks negotiation. Raising your own voice, which is the most compelling reaction of all. Caving on the original limit to end the noise.
Expect this: An extinction burst -- a period where the screaming gets louder and more intense before it reduces, because the tool that used to work is not working and the child is applying more force to it. This is the hardest moment and also the clearest sign that the change in your response is being registered. Hold through it.
This one has a particular quality that parents recognise immediately once it's named: the screaming is too aware. There is something in it that feels directed. The child is watching while they scream. The volume adjusts depending on whether you're in the room. It stops, sometimes completely, the moment something genuinely interesting happens. And it started, very precisely, the moment you did something they didn't want.
A child who screams to get a reaction has discovered one of the most reliable tools available to a young child: that a particular sound, produced at sufficient volume, reliably causes a specific person to change what they're doing. This is not manipulation in the adult sense of the word. It is cause-and-effect learning, which is exactly what children this age are wired to do. They found something that works. They are using it.
What makes this hard is that the reaction they're looking for is not always obvious. It isn't only about getting the thing they want -- the screen back, the snack, the different outcome. Sometimes the reaction they're seeking is simpler and more primal than that: intensity. A parent who raises their voice, or whose face changes dramatically, or who rushes across the room to intervene is providing something the screaming was built to produce. The child may not have wanted to be shouted at. But the shouting was a response, and a response is what they were looking for.
The screaming that is working is not evidence of a difficult child. It is evidence of a child who has run an experiment, observed the results, and drawn an accurate conclusion. The conclusion needs to change.
The screaming is clearly reactive -- it starts at limits and transitions, it has a watchful quality, and it reduces when you don't engage with it. Your child is between two and four. There are also many moments in the day where they manage disappointment without screaming. The pattern is frustrating but not all-consuming.
Worth watching if the screaming is happening across all contexts regardless of whether a limit was set, the intensity is escalating rather than stabilising, or the pattern is significantly affecting daily life -- meals, outings, sibling relationships. A child who screams to communicate more often than they use any other tool needs more support than a change in parental response alone.
The screaming is accompanied by breath-holding that results in passing out, significant self-harm, or a level of dysregulation that suggests the behaviour has tipped past instrumental and into something the child themselves cannot control.
The reaction is keeping it going. Specifically, the inconsistency of the reaction.
A parent who responds neutrally most of the time but breaks at the fifth or sixth minute -- offering something, raising their voice, or caving on the original limit -- has not failed to respond consistently. They have responded very consistently: they have taught their child that the screaming needs to run for approximately five minutes before it produces a result. The child is not being unreasonable. They are being accurate.
Intermittent reinforcement -- the pattern where a behaviour is sometimes rewarded and sometimes not -- produces the most persistent behaviour of all. A child who gets a result every time they scream will stop screaming the moment the results stop. A child who gets a result one time in six will scream for much longer before concluding it isn't working, because last time it worked on the sixth try.
The second thing that keeps it going is the quality of the reaction when it comes. A parent who holds for five minutes and then shouts is providing a more intense reaction than a parent who responded calmly at minute one. The screaming has been building toward something, and the explosion at the end of the hold is a bigger result than the calm engagement at the start would have been. From the outside this looks like the screaming caused the shouting. From the child's nervous system, the shouting was the point.
The third thing: the absence of a replacement. A child who is told that screaming doesn't work needs something that does work. A clear, accessible alternative -- using a word, a sign, a physical signal, anything that produces a reliable response -- needs to be available and practised before screaming is the only tool that has ever gotten results.
