Parts integration is the NLP technique for inner conflict, the “part of me wants to go, part of me wants to stay” state that no amount of thinking resolves. You separate the two competing parts, put one on each hand, ask each of them what it is ultimately trying to do for you, keep asking until both arrive at the same higher intention, and then bring them back together. A full run takes twenty to forty minutes, and it works on exactly the conflicts that produce the same stalemate every time you revisit them.
| Stage | What you do | What you are listening for |
|---|---|---|
| Separate | Put each part on one hand, describe how it looks and sounds | Two clearly different voices, not one voice arguing |
| Chunk up | Ask each part what it wants, then what that gets you | Answers moving from behaviour to purpose |
| Meet | Keep chunking until both give the same answer | A shared intention, usually one word |
| Integrate | Let the hands come together at their own pace | A physical change: the hands move without effort |
| Check | Test the old trigger situation in your mind | Whether the argument still starts |
What counts as an inner conflict?
An inner conflict is a decision that keeps reopening. You resolve it on Tuesday, and by Thursday the other side is talking again. The giveaway is the phrasing: “part of me wants to hand in my notice, but part of me thinks I would regret it.” People say that sentence without noticing they have just described two separate things inside one head.
That is different from indecision about facts. If you cannot choose between two flats because you have not seen the second one yet, you need information, not a technique. Parts integration is for the case where you already know everything you are going to know and the loop continues anyway. The reference entry on what parts integration is and where it came from covers the model; this article is the drill.
The idea that a person contains parts with their own agendas is older than NLP. Fritz Perls worked it with two chairs in Gestalt therapy long before Richard Bandler and John Grinder described the hands version, the “visual squash”, in Frogs into Princes (1979). What NLP added was the chunking sequence: the assumption that two parts in conflict are both working for you, at cross purposes, and that if you ask far enough up you reach the point where they agree.
How do you run parts integration step by step?
Sit somewhere you will not be interrupted, with your hands free and palms up on your knees. Work out loud if you are alone. Saying it aloud keeps the two parts from merging back into ordinary internal chatter.
- Name the conflict in one sentence. “Part of me wants X, and part of me wants Y.” If you cannot say it in that form, you are probably not dealing with two parts yet.
- Put the first part on one hand. Let it sit on the palm. Give it a shape, a colour, a size, a voice. This sounds odd and it is the step that makes the rest work, because a part you can see and hear is a part you can talk to.
- Put the second part on the other hand. Same treatment. Notice how different the two are. They usually feel different in weight and temperature, and that difference is useful information.
- Ask part one what it wants for you. Not what it wants you to do. What it wants for you. Then ask: and if it had that, fully, what would that get you that is even more important?
- Keep going up. Ask that question five or six times. The answers climb from actions (“leave the job”) to states (“stop feeling trapped”) to something close to a value (“be free”).
- Do the same with part two. Start at the top of its own chain and climb the same way. Do not compare the two while you climb. You will be tempted to.
- Find where they meet. At some level, both give you the same answer, or two answers that mean the same thing. Say it out loud. This is the moment the conflict stops being a conflict, because both sides turn out to have been chasing one thing.
- Let the hands come together. Do not push them. Ask each part whether it would be willing to have the other one’s resources, since they want the same outcome anyway, and then let your hands move at whatever speed they move. Bring them to your chest when they meet.
- Test it. Think forward to the next situation that used to trigger the argument. Notice whether it starts.
Step eight is the one people describe afterwards as strange. The hands often move slowly, unevenly, and without a decision to move them. That is normal. If they do not move at all, you almost certainly have not found the shared intention yet, so go back to step seven.
What do you ask to find the higher intention?
One question, repeated: “and what does that get you that is even more important?” Everything else is a variation. The value is in the repetition, not in clever phrasing, because each repetition strips one more layer of behaviour off and leaves purpose behind.
- If the answer is a behaviour (“hand in my notice”), ask what having that would do for you.
- If the answer is a negation (“not be exhausted”), ask what you would have instead.
- If the answer repeats, you have probably reached the top of that chain. Ask once more, and if it repeats again, stay there.
- If the answer is about someone else (“so my family is fine”), ask what that would give you personally. Parts work for you, not for other people, even when they are protecting a relationship.
The two chains meet more often than beginners expect. A part pushing you to leave and a part pushing you to stay both turn out to want safety, or self-respect, or a life that feels like yours. They disagreed about the route, never about the destination. That is the insight the technique is built on, and it is also why arguing with yourself never settled it: both sides were right about the goal.
Before you finish, run an ecology check on the change you are about to make. Ask whether anything in your life depends on the conflict staying unresolved, because occasionally it does, and that is worth knowing before the hands come together rather than after.
When does parts integration not work?
Four situations, and knowing them in advance saves you from concluding the technique is nonsense.
When there are more than two parts. Some conflicts have three or four voices. Running it as a two-hand technique then produces a merge that unravels within a week. Sort out which two are actually in opposition first, or integrate in pairs.
When one part is a stand-in for a person. If “the part that wants me to stay” is really your mother’s voice, you are not integrating two parts of yourself. Perceptual positions handles that better, because the problem is a relationship rather than an internal split.
When you stop at pros and cons. Listing what each side wants is not the technique. Without the climb to a shared intention there is nothing to integrate, and you finish with a tidier version of the same argument.
When the conflict sits on top of something clinical. Inner conflict that comes with persistent low mood, intrusive memories or a sense of parts that feel genuinely separate from you belongs with a qualified professional. NLP work of this kind complements professional care and does not replace it, and a trained practitioner should recognise the difference and say so.
How is it different from weighing up pros and cons?
A pros and cons list treats the two sides as arguments to be scored. Parts integration treats them as two intentions to be reconciled, which is why one produces a decision you then have to defend against yourself, and the other produces a decision that stops being contested.
There is also a structural difference. A list keeps both sides at the level of behaviour, where they genuinely are incompatible: you cannot both leave and stay. The chunking step moves them to a level where compatibility is possible, because “freedom” and “security” can coexist in a way that “quit” and “stay” cannot. This is the same upward move described in the entry on chunking up, down and sideways, applied to two things at once.
If the conflict is really about how you frame one event rather than two competing intentions, the six-step reframe is usually the better tool. And if you would rather run this on something live with someone watching, it is a standard piece of one-to-one NLP coaching work, in Lisbon or online.
What I notice when people run this in a room
The tell is in the hands, and it arrives before anyone says anything. People start the exercise holding both palms flat and symmetrical, like a polite gesture. Somewhere around the third or fourth chunking question one hand starts to sit differently from the other, higher, or cupped, or slightly turned away. That asymmetry means the two parts have genuinely separated, and until it shows up I know we are still working with one voice pretending to be two.
The other thing I have learned to watch for is the rush at the end. Once someone hears the shared intention they want to clap their hands together and be finished, and a forced merge does not hold. I ask them to wait and let the hands do it. In a group of ten there is usually one person whose hands take a full minute, and they are almost always the one who reports the biggest change the next morning.
Carolin Mallmann, Licensed Trainer of The Society of NLP, Lisbon
Frequently asked questions about parts integration
How long does parts integration take?
Twenty to forty minutes for a first run on one conflict. Most of that time goes into the chunking questions, which need to be asked slowly enough for a real answer to arrive. Later runs on smaller conflicts can take ten minutes once the sequence is familiar.
Can I do parts integration on myself?
Yes, and it is one of the NLP techniques that survives solo use reasonably well, because the structure carries it. Working out loud helps. A partner is useful mainly for noticing when you have answered a chunking question with a new behaviour instead of a purpose, which is the most common way a solo run stalls.
What if the two parts never agree?
They usually do if you keep chunking, because at a high enough level almost every part wants something like safety, belonging or self-respect. If you have asked six or seven times and the answers still differ, check whether one of your parts is actually two parts, or whether one of them represents another person rather than you.
Is parts integration the same as the visual squash?
They are closely related and the names are often used interchangeably. The visual squash, described by Bandler and Grinder in the 1970s, is the hands-and-images version. Parts integration is the broader label for the same process, usually taught with the chunking sequence to a shared intention made explicit.
How do I know whether it worked?
Test it on the trigger. Imagine the next situation that used to start the internal argument and notice whether it starts. A run that worked tends to feel unremarkable afterwards, because the question simply stops being interesting, rather than dramatic. Check again a week later.
Keep exploring
- The six-step reframe, step by step, the other NLP technique that negotiates with a part rather than overruling it.
- The perceptual positions exercise, published this week, for conflicts that turn out to involve another person rather than two parts of you.
Carolin Mallmann is a Licensed Trainer of The Society of NLP, trained by Richard Bandler. She has trained more than 300 people and runs English-language NLP trainings in Lisbon, in groups of no more than ten.
Learn parts integration with someone watching your hands, on the 7-day Practitioner training.
