Select Page

Chunking: what is the meaning? [All questions!]

Chunking: what is the meaning? [All questions!]

What is Chunking? In this article you will find the meaning, examples and questions you can ask for upchunking or downchunking . This is a powerful method from psychology nad NLP. Read along…

What is chunking?

The idea of ​​chunking is inspired by the work of Bertrand Russel and Gregory Bateson. Chunken is a way of creating a hierarchy in how specific or abstract you communicate something.

  • When you upchunk, you move up the hierarchy of abstraction. For example, “transport” is a higher logical level than “train.”
  • When you downchunk, you go down the hierarchy of abstraction: you become more specific. For example: ‘train’ is an example of transport. If we downchunk further, an intercity is an example of a train.

We organize our experience in unconscious hierarchies. The function of each level is to organize the information at the level below it. These levels that you can up or down to are called logic levels (not to be confused with Bateson’s and Dilts’ six logic levels ).

How can you chunk? 3 ways!


In the next three paragraphs, let’s zoom in on the three possible ways of chunking.

Way 1 is upchunking: to a higher level of abstraction

By upchunking, you become more abstract, so that you get further and further out of the problem. The linguistic result is basically the ambiguous language of the milton model.

Is there still a problem? Then there is not yet high enough.

For example:

  • What is this an example of?
  • For what purpose?
  • What is your intention?
  • What is the positive intention?
  • What’s important to you?
  • What do you really want?
  • What do you think is really important?
  • For what purpose?
  • To achieve what?
  • To get where?
  • What is this part of?
  • What does that do for you?
  • What does it want for you?
  • How does it serve you?
  • What does it mean to you?
  • What does it bring you?
  • What does it give you?
  • And if you do, what’s in it for you?
  • What’s in it for you? What would that do for you?
  • What benefits do you want to get from it?
  • How does that make you feel?
  • What does that lead to?
  • What is important to you in this respect?
  • What’s even more important?
  • And if that happens, what are the effects of that?
  • And the result was?
  • What is the impact of that?
  • What motivates you to do x, to do what you do?
  • When you reach x, what happens? (Keep asking this question over and over to go deeper and deeper.)
  • What are you doing it for?
  • What is the reason you are doing this?
  • So you…
  • Which makes you …
  • The Buddha Technique: Why …? Why…? Why…? Continue in absolute silence and attention.
    (Upchunk questions have the property of making a person feel associated. In addition, the last question is specifically designed to keep the person extra associated.)
  • You can find more upchunking in this article about the logic levels of Bateson & Dilts.

An example of upchunking along different logical levels of learning:

  1. ‘Learning’ is at a certain level.
  2. Higher level: learn how to learn
  3. Even higher level: learning how you learned how to learn.

Let’s dig a little deeper into this example.

  1. I can learn how to add, subtract and multiply without understanding how they work.
  2. At the next level I learn to make my own calculations and find out how I divide something by something.
  3. At the next level, I learn how I learned the process of making these calculations, and I move straight to the next level, opening up many more learning opportunities in the lower sets.

Way 2 is downchunking: to a lower abstraction level

You go for details, specifications and data. This is the specific language of the meta-model.

For example:

  • What are examples of this?
  • What specifically?
  • How specific?
  • How do you know that?
  • All other questions of the meta-model.

Use downchunking, for example, when giving and receiving feedback : address behavior instead of identity. If you are being accused yourself, post the problem with a gesture in front of you where you are pointing and ask questions about the lower levels: “What exactly happened, I want to know everything.” In this way you go into behavior and environment instead of your identity.

Examples of downchunking:

  • “You’re mean.”
    “What part of my sentence do I have to change to stop being mean?”
  • “The US is a rich country”
    “Is a resident of the US automatically rich?”
  • An NLP person often asks: “How do you do that?” This makes something more specific. This releases all kinds of information, for example about the meta programs and the representation system of the other person: “I know that I am good at my job. I understand my profession very well.” “How do you know that?” “I talk to the end users, I ask them questions and I hear them say that they …”

Way 3 is lateral chunking: first chunk up and then chunk down

With lateral chunking you create similar alternatives. So this is ideal for troubleshooting. Lateral chunking is typically something Steve Jobs did. Some quick examples of questions for chunking laterally:

  • Is there anything like this?
  • Is there anything equal to this?

The steps of lateral chunking:

  1. Chunk up.
  2. Chunk down: what are other examples of this?

Applying lateral chunking in practice: that’s how valuable this reframing method is


Lateral chunking is a form of reframing You reframe to higher or lower logical levels to find new alternatives. You also treat the positive intention of a negative behavior or feeling by means of upchunking and then downchunking.

Let’s take an example. If someone wants something from you, you can ask, “Why do you need it?” Then devise another method to achieve the same. Give him that, and hopefully it won’t be as bad as his first demand. You are first upchunking to the goal, and then downchunking to the means to achieve the goal.

Note: Those new alternatives may be more difficult and less out-of-the-box to get to than the easy quick way. For example: cigarettes or chocolate provide immediate and effortless pleasure, while for other forms of pleasure, such as painting or a massage, you first have to set up all kinds of things and make appointments.

More examples and applications of lateral chunking

  • “What benefits do you get from…” “Safety.” “How could safety help? So what does safety ensure? ” “Peace and tranquility.” “So what you really want is peace and tranquility.” Now you can downchunk to behavior again to find other options for peace and tranquility.
  • Problem: We cannot travel by car.
    Upchunk: what is the purpose of a car? Transport.
    Then downchunk: what are several forms of transport?
  • Upchunk in arguments to discover which values ​​are not being fulfilled. Then find new ways to fulfill those values ​​through downchunking.
  • In modeling , three musketeers don’t fit in my office when I visualize them, but what is their quality? They are brave. Who else is brave? Oprah.
  • “He thinks I’m untrustworthy …” “Why is it important for you to be trusted?” “Self-protection.” Now opportunities are emerging to get started with self-protection and to notice that it is so nice to have, so that the problem of being found unreliable is no longer important.
  • “What did you mean to achieve with …” instead of, “That was ridiculous.”
  • “The US is a rich country”
    First, a downchunk example:
    “Is a US resident automatically rich?”
    Lateral-chunk example:
    “That would be true if money were the sole criterion for wealth. In what other ways can you be rich? ”

Chunking and NLP techniques

upchuncken downchunken

Chunking plays a major role in NLP techniques: the most important reframing, the intention reframing, is an example of this. In six-step reframing  , you are busy upchunking to the positive intention, and then downchunking again to new examples of that positive intention.

Chunking and logical levels also play a role in the NLP technique Visual Squash . It is important to know that not all parts should integrate with each other. They are then different logical types. In that case you do a parts negotiation instead of a parts integration. Can they coexist? What should it take for that? For example a bridge?

About The Author


Hello! Thanks for reading these articles. My intention is to make happiness as simple and clear as posssible. By the way, excuse my English. I am not a native English speaker since I live in Amsterdam. Much appreciated if you use the comments to make suggestions on my grammar. See ya in another blogpost!