Certainty and Uncertainty*
by Steve Andreas
©2000 

       The following is a very interesting exchange between Richard Bandler and someone who is very sure about something.

B: Are you sure?
P: Yes.
B: Are you sure you're sure?
P: Yes.
B: Are you sure enough to be UNSURE?
P: Yes.
B: OK, Let's talk.

         Before reading further, I strongly recommend that you think of  something that you are very certain about, and find someone else to ask you this set of questions about your certainty, so that you have a  concrete personal experience of their impact. At the very least, close your eyes and imagine that someone else asks you these questions, and take the time to carefully notice your response to each one, so that you can experience their effect on you.
         And for those of you who teach modeling, or do modeling, this is an excellent small opportunity to do some of it. Although Bandler’s exchange is brief, and concise, it is quite interesting to explore its structure.                                                 

                                             *       *       *

         Now that you have an experience of it, I would like to characterize this pattern as I understand it, which requires a short journey up through logical levels.
Level 1.
         There is a situation X. X is an event in more or less sensory-based, "reality," what Paul Watzlawick has called "first-order reality." This is something that everyone can usually pretty much agree on, such as a "job interview," or a "critical comment." This level is often called "the environment," and it is something that often we don't have too much control over. Certain unpleasant events  happen to us from time to time, and we don't always have the choice of avoiding them or ignoring them.
Level 2.
         The person then thinks about the situation X in a particular way and characterizes/evaluates it, for instance, "This X is scary." This is a meta-response, and the state is a meta-state about  X. This is what Paul Watzlawick has called "second-order reality." This is where people may differ wildly, particularly if they are from different cultures, and it is at this level where many conflicts and problems (and many solutions) exist.
         The person could just as well conclude that X is "boring" or "exciting," or "challenging," or is an opportunity to "learn more about their Buddha nature," etc. The person’s response will depend on the understanding that they apply to the event, and changing this understanding through content reframing can make a huge difference in the person’s experience.
Level 3.
         The person has a degree of certainty about the meta-response. "I know  this is scary." This is a meta-response about a meta-response (a meta-meta-response, with corresponding meta-meta-state). We could call this "third-order reality," which is even more distant from sensory experience than  second-order reality, and even more troublesome and dangerous. Plenty of problems (and solutions) also occur at this level.
         Many people who come for therapy appear to suffer from uncertainty: "I don't know what to do." "I'm not sure if this is the right thing to do." "Life has no meaning." But you can also think of this as resulting from other certainties. "I know that wouldn't work," "I know  she hates me," "I know  I can't succeed," etc. Since these certainties will make it difficult for the person to consider other understandings at level 2, it can often be very useful to reduce certainty.
         Someone who is phobic of airplanes, and someone who is not, may be making exactly the same images of flaming death and destruction. The difference is that the images of the non-phobic include some representation of the small probability  of the crash, as well as its possibility.  This could be either a certainty of its unlikeliness, or a very great uncertainty about its happening. However, a phobic person is experientially certain that it will happen, no matter what s/he says "intellectually."
         What makes it difficult to work with a paranoid is not just that s/he thinks that others are plotting against him/her, but that s/he is certain that this is occuring, and is unwilling to question it and consider other possibilities.
         Another aspect of a person’s certainty is that others may suffer from it as much or more than the person who is certain. Think of all the deaths, persecutions, misery and destruction around the globe that have resulted from the certainty of religious prophets and institutions, revolutionaries, and politicians--all of whom are totally convinced that they were right.
         Each of us has a way to assess experience and provide us with a measure of how certain we are about it. This has often been called a person’s "convincer strategy." The exploration of the variety of ways that people use to convince themselves of something is also relevant to the topic of certainty, but this article will only discuss the result of the operation of the convincer strategy.
         Every  evaluation that someone makes at level 2 has some degree of certainty/uncertainty about it at level 3, and this will be on a continuum from zero certainty to  absolute certainty. There are basically three possibilities:
A. Zero certainty
         If a person has zero certainty, they have no firm conclusion whatsoever about the meaning of X, so they are completely open to considering new understandings when they are offered, and they will be very easy to work with in exploring other ways of thinking about the situation X. This is an "easy client," because their understanding of a situation is very fluid, and they have no, or very little, certainty about their understanding to lock in the understanding, and make it hard to change. 
B. Partial certainty
         If someone is somewhere in the mid-range of certainty, they are at least somewhat open to considering other possible understandings (on level 2) of a situation X (on level 1). If they are very certain, it will be harder for them to consider other understandings, but at least it will be possible. These clients are  somewhat harder to work with than those with zero or very little certainty, and those who are more certain will be harder to work with than those who are less certain.
C. Absolute certainty
         If a person is totally certain about their understanding, they will be closed to even considering other understandings, because their certainty about their understanding locks up the ability to consider alternatives. These are the really tough clients, and this is the situation where Bandler's pattern is particularly useful--to move someone from the absolute certainty (which has only one representation) to the partial certainty (with more than one representation) in which a dialogue is possible. (I think it is very significant in this regard that at the end of the exchange, Bandler says "OK, Let's talk.") In other words, this pattern is not useful to solve a problem, it is useful to make it possible to solve a problem on level 2 by decreasing certainty on level 3.
Understanding the pattern
         To understand how the pattern works, we will need to enter the realm of paradox, which is very difficult for most of us to think about. (It was also hard for Bertrand Russell and Gottlob Frege, two very brilliant professional logicians to think about, so there is no shame in this, but the faint of heart may wish to consider turning to simpler recreations.)
         "Are you sure?" asks if a person is in state of certainty. This is a question that asks for a digital yes or no answer, but permits answers which are qualified in some way.
         If the person says, "No, not really," then they are uncertain (A) and are already open to other understandings.
         If they respond, "Well, I'm pretty sure," they are somewhere in the intermediate range of partial certainty (B) and will be at least somewhat open to considering other understandings.
         If they simply respond "Yes," we need more information. (As usual the nonverbal messages in voice tone, posture, hesitations, etc.  will be much more useful than the words in assessing the actual degree of certainty the person is experiencing.)
         "Are you sure you're sure?" applies certainty to itself recursively, in essence asking if the person is absolutely sure. Answering this question requires the person to go to a 4th level, applying certainty to itself. Again this is a question that asks for a digital yes or no answer, but permits a qualified answer.
         If the person says. "Well, I'm pretty sure," or qualifies it in any way, then the person is somewhere in the mid-range (B), and can  already be talked with usefully. If the person replies with an unqualified "Yes," they are saying that they are absolutely certain (C). (Again, the nonverbals will tell you more about the absoluteness of the certainty than the words.)
         This condition of absoluteness (or near absoluteness) is required for the next step of the pattern to work. However, if the condition of absoluteness is not met, it means that the next step is unecessary, because in a condition of partial certainty (B) you can proceed to usefully explore alternative understandings.
         A very important aspect of this question is that it asks the person to recursively apply their certainty to itself. This requires the person to go to a fourth logical level, and this is something which is also necessary for the next step in the pattern.  A "Yes" answer is a confirmation that the person is willing and able to do this recursion or "apply to self," as it is usually called in the "sleight of mouth" patterns.  Recursion is a precondition for the next question, which also asks the person to apply certainty to itself, but in a different way.
         Another way of describing this is that the first two questions can be used both to gather information about the client's degree of certainty, while at the same time beginning to assemble pieces of a puzzle which will be put all together in the third step.
        "Are you sure enough to be UNSURE?"  applies certainty to its negation, and is a form of logical paradox, equivalent to the statement "This sentence is false (not true)," or "I am a liar (not truth-telling)." (The word "paradox" can also used in a more general way to mean "contradictory" or "unexpected," but the meaning here is restricted to  logical paradox.)

The three essential ingredients of a logical paradox are:
       1.  An absolute statement,
        2.  Recursion, 
        3.  Negation.

         In paradox, an absolute statement is recursively applied to its own negation, bridging two logical levels. If  the statement is true, then it is false, and if it is false, then it is true. This perpetual oscillation between truth and falsity challenges all our ideas about certainty and reality, and this is at least one reason why we find it so difficult to think about paradox.    
         There are two more very important elements in the word "enough." "Enough" presupposes some point on a continuum, while the person has been using an absolute either/or (sure/unsure) distinction with no middle ground. No matter how the person answers, if they accept this presupposition, they are agreeing to a frame in which certainty is on an analog continuum rather than an absolute, digital either/or, and consequently other alternative understandings can be considered.  Unless they challenge this presupposition, either answer to this question moves them to an experience of partial uncertainty.
         There is yet another important element in the word "enough." It presupposes reaching a threshold, in this case a threshold of certainty. If the person replies "No," they are saying that their certainty is something less than the threshold. If they reply "Yes," they are saying that their certainty has reached (or exceeded) the threshold, and is "enough" to be uncertain.
         "Are you sure enough to be unsure?"  is the question form of the statement, "If you are sure enough, you will be unsure," and this is presupposed when asked as a question. This presupposition states that great certainty includes within it the ability to be unsure, taking two experiences that have been experienced as polar opposites, and nesting one within the other.
         I have already mentioned that it is very difficult for most of us to process logical paradoxes. When we hear this paradox, stated as a question, (with the "enough" presuppositions packed inside it), most people simply give up and respond yes or no.
         If a person answers "Yes," they are agreeing to a state of unsureness (the "unsure"), and if they answer "No," they are also agreeing to a state of unsureness "not sure enough." Whichever response is given, they are agreeing to a degree of uncertainty, and consequently the willingness to consider alternative understandings.           
         This pattern has the same form as a paradoxical challenge that the devil supposedly once offered to God in regard to God's omnipotence. The devil challenged God to create a rock so large that even God could not move it. If God cannot create a very large rock that he cannot move, he is not omniptent in his ability to create rocks, and if he does create such a rock, he is not omnipotent in his ability to move rocks. Either way the absoluteness of God’s omnipotence is destroyed.
         To summarize, this pattern is very useful in situations in which a person is very certain about how they understand something, this understanding causes them difficulty, and their certainty results in their being not willing to even consider alternative understandings. Using this pattern can open them to considering other models of the world.
         Learning how to sort out levels of experience in this way is a very useful skill that can help us understand the structure of problems, and decide which level of understanding could use some improvement.
         This makes it much easier to find our way through the twisting corridors of another person's mind, in order to help them find their way out of their predicaments--and also keeps us from wasting our time solving problems that they don't have!
         Confusion about levels of thinking, the recursion which transcends levels, and particularly recursion that includes negation, are present in many human problems. It is a little-explored realm, and one that often creates paradoxical traps for us. Knowing the three essential elements of paradox (absolute statement, recursion, and negation) can help us identify these traps, and avoid them.
         We can't avoid logical levels, or recursion, and we wouldn't want to--that would keep us from thinking about thinking, and having feelings about feelings, thinking about feelings, and many other valuable and unique aspects of our humanity.
         But we can learn to use positive statements whenever possible, rather than negations, and learn to be very  careful when we do use negation. The NLP emphasis on positive outcomes is one example of the value of this, and the benefits that can result from this kind of care in thinking.
         And we can be doubly careful when recursion is also present, which is much more often than we usually think. To give only one example, when someone says, "I am a bad person," they are saying that everything that they do is bad, and one of their behaviors is the sentence that s/he just said to you, so "badness" applies to the sentence about badness.
         And finally, we can also learn to be very cautious about making absolute statements, realizing that all knowledge is relative, contextual, and based on our very limited experience and understanding.  Paradoxically, that is one thing we can be very certain about!
         I think it is truly amazing that with the three pounds of jelly between our ears we can imagine and think about an infinite universe, but it would be useful to have a little humility all the same. Let’s start with some humility about our knowledge and certainty.
         In case the reader at this point is still insistent that there is such a thing as absolute certainty, I offer the following quote from Warren S. McCulloch’s 1945 article "Why the Mind is in the Head," now included in his marvelous book Embodiments of Mind, MIT Press, 1965.
         McCulloch was one of the first and the best to apply mathematical analysis to the functioning of the nervous system.
         "Accordingly to increase certainty, every lhypothesis should be of minimum logical, or a priori, probability, so that if it be confirmed in experiment, then it shall be because he world is so constructed. Unfortunately for those who quest absolute certainty, a hypothesis of zero logical probability is a contradiction, and hence can never be confirmed. Its neurological equivalent would be a neuron that required infinite coincidence to trip it. This, in a finite world, is the same as though it had no afferents. It never fires."

*Anchor Point, Vol. 14, No. 10, October, pp. 3-8