The question arrives fully formed and it contains a premise that is doing all the damage. Someone was told, they went ahead anyway, and now the search is for the thing to say that will make the telling work next time.
There is no such sentence. If a limit only holds when the other person agrees to hold it, it was never a limit. It was a request, and it was granted for a while.
46%
of American workers with paid leave take less time off than they are offered
49%
of them say it is because they would fall behind
43%
say they feel bad about colleagues absorbing the work
5,188
workers in the survey
A limit that needs their cooperation was a request
Requests are not a lesser thing. Most of what keeps any relationship running is requests, made and granted, without ceremony. The problem is what sits underneath one when it is ignored, which is nothing at all, and that emptiness is what brings people to this question.
The entry on personal boundaries makes the distinction in one line: a boundary is a statement about what you will do, not a rule for what someone else must do. Stop checking my phone is a rule. If my phone gets checked, I am going to change the passcode is a boundary. The second does not need agreement and does not improve if it is phrased more assertively.
That is the whole of the conceptual half, and most articles stop there, which is where they become useless. Everyone who has read anything about this already knows the distinction. What they do not have is an account of why the second sentence still fails in the moment, and there is a good one.
The gap between deciding and doing has been measured
Webb and Sheeran pooled 47 experimental tests of whether changing what people intend to do actually changes what they do.
- Change in intention: 0.66d
- Change in behaviour: 0.36d
A medium-to-large shift in intention produced a small-to-medium shift in behaviour. Roughly half the change evaporates somewhere between deciding and acting, and this is in controlled experiments on ordinary goals with no emotional history attached to them.
Now put a person in a kitchen at eleven at night, in a relationship they care about, facing someone who has just crossed the line for the third time and is looking slightly hurt about being told. The intention was formed weeks ago in a calm hour. The moment it has to survive is not calm.
So the failure is not weakness of character and it is not a phrasing problem. It is the ordinary intention-behaviour gap, arriving under load, in the one domain where the cost of acting is a person being upset with you.
The form that closes it
Here the research is unusually practical.
Gollwitzer and Sheeran's meta-analysis covered 94 independent tests and more than 8,000 participants, comparing goals held as general intentions against the same goals specified as if-then plans. The if-then version produced a medium to large effect on attainment, around .65. The difference between the two conditions is nothing but grammar: a plan that names a trigger and an action, rather than a resolve that names an outcome.
I will stop letting this happen is a goal. When the conversation goes past eleven, I am going to say goodnight and go to bed is an implementation intention, and the second one survives the eleven o'clock kitchen for a mechanical reason. The situation itself does the remembering. Nothing has to be decided at the moment when deciding is hardest.
Which is why the sentence the boundaries literature has recommended for decades works better than the reasoning behind it usually suggests. When X happens, I am going to Y is not a communication template. It is the form of plan that has the best evidence behind it for getting anything done at all, and boundaries are simply the version of it where the obstacle happens to be someone's feelings.
| Version | What it depends on | What happens at the first crossing |
|---|---|---|
| Stop doing that | Their agreement | Negotiation, and the limit is now a topic |
| I need you to respect my time | Their interpretation of a value | Agreement in principle, no change in practice |
| I would really rather you did not | Their willingness to read a hint | Nothing, usually, and resentment later |
| When X happens, I am going to Y | Nobody but you | The action, which is the actual boundary |
The leave nobody takes
There is an everyday proof of all this that has nothing to do with relationships.
Pew surveyed 5,188 American workers in 2023 and found that 46 percent of those with paid time off take less of it than they are entitled to. The entitlement is written down. It is contractual, uncontested and requires no negotiation with anyone. And nearly half of the people who have it decline part of it.
The reasons they gave are the interesting part. A slight majority, 52 percent, simply did not feel they needed more time off, which is a real answer and probably a true one. Of the rest, 49 percent worried about falling behind at work and 43 percent felt bad about colleagues absorbing what they left.
Set the 52 percent aside and what remains is a clean demonstration: a limit that has been formally granted, in writing, by an authority that will not contest it, and it still does not get taken, because taking it is an action and the action costs something socially. Nobody is ignoring these boundaries. There is no antagonist in the story at all.
Where this framing breaks down
Two places, and they matter enough that the rest of the piece is dishonest without them.
The first is asymmetric power. All of this assumes an action is available at an acceptable price. With a manager, a landlord, an immigration sponsor or a parent who still controls the housing, the action is available in theory and the price is the thing that makes it unavailable, and advice that skips over the price is advice for people who are not in that situation. What remains genuinely useful there is narrower: smaller actions, documentation, and the honest recognition that this is a constraint problem rather than a skills problem.
The second is that a person who crosses every limit, regardless of how it is framed, is not failing to understand. Repeated deliberate crossing is a description of the relationship, not a problem with the sentence, and the reading that matters at that point is toxic relationships rather than anything about technique. The difference is visible in what happens on the third occasion rather than the first. Initial pushback is ordinary. A reaction so reliably large that limits stop being stated at all is the pattern itself.
The version of this question that has an answer is not what do I say when they ignore it. It is what do I do when they ignore it, and that one has been answerable the whole time. It was just a worse answer than the one people were hoping for.
Sources
- Webb, T. L., Sheeran, P. (2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132(2).
- Gollwitzer, P. M., Sheeran, P. (2006). Implementation intentions and goal achievement: a meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38.
- Pew Research Center (2023). About 46% of US workers who get paid time off do not take all of it.
- Gollwitzer, P. M. (1999). Implementation intentions: strong effects of simple plans. American Psychologist, 54(7).
- Sheeran, P., Webb, T. L. (2016). The intention-behavior gap. Social and Personality Psychology Compass, 10(9).