A.N.Z.J. Fam. Ther., 1997, Vol. 18, No.

2, pp 109114

EDUCATION UPDATE:
Fundamentals of Theory and Practice Revisited

Circular Questioning: An Introductory Guide*

Jac Brown**

This paper presents a simplified model that has been found to be useful for those learning questioning
skills within the Milan systemic approach to family therapy as well as for those who simply wish to develop
some skills in circular interviewing. The model begins with the premise that clients frequently describe a
problem in terms that are either too broad or too narrow. The model has two categories of questions: those
that draw connections and thus broaden a clients understanding of their context and those that draw
distinctions and thus narrow a clients focus where generalisations predominate. The model is presented
within the context of other models of Milan systemic questioning and the contributions they have made.

INTRODUCTION

connections or distinctions in the surrounding environment. For example, if I state that she is happy, I am
stating this in relation to other people I have seen who
were less happy. Thus, the noted difference in happiness has information value. The questioning is aimed at
creating or maximising difference and then drawing
connections in order to provide information that frames
problems in new ways. A number of writers have subsequently devoted considerable effort to the task of delineating uses for circular questioning, in the process of
facilitating the release into the system of new information about the problem and its context (Penn, 1982;
Fleuridas, Nelson and Rosenthal, 1986; Tomm, 1987;
MacKinnon, 1988). This paper examines a two factor
model of questioning which aims to help therapists to
broaden the familys focus when it is too narrow, and
narrow the focus when it is too general. It is particularly
useful as a framework for family therapy students
attempting to construct their own circular questions.

The interviewing style of the Milan Associates has provided a unique contribution to the field of psychotherapy: the process of circular questioning. Circular
questioning has since been adapted to many models of
therapy. The development of circular questioning
emerged from the Milan Associates application of Batesons ideas about circularity to their behaviour as therapists. They state that their research was directed at . . .
the most correct and fruitful procedure for interviewing
the family (Selvini Palazzoli, Boscolo, Cecchin and
Prata, 1980: 3). Circular questioning draws connections
and distinctions between family members or people
within the larger client system. For example, the behaviour of one person is shown by implication to be connected to the behaviour of another in a circular manner
rather than in the usual lineal or causal way that has
been the basis of much of our thinking about human
problems. Thus, instead of asking why someone is
depressed, a circular form of questioning would inquire
about when someone shows depression and what other
people do when this is happening.
This interviewing style stimulates the release of information into the system in a manner that encourages new
ways of viewing the problem. The premise behind circular questioning is that information comes from difference and that difference implies a relationship, through

THE DEVELOPMENT AND ELABORATION

OF CIRCULAR QUESTIONING
The Milan Associates do not say a great deal about actual
interviewing in their seminal 1980 paper (Selvini Palazzoli, Boscolo, Cecchin and Prata, 1980). However, what
they do say is extremely significant and has stimulated
extensive thinking and subsequent delineation. Citing
Bateson, they work within the framework that information is a difference and that difference is a relationship or a change in the relationship. They highlight the
benefits of investigating a dyadic relationship by asking
a third person for his or her perceptions on that other

* The author wishes to thank Kerrie James, Director of Clinical

Services at Relationships Australia (NSW) for her helpful comments.
** Jac Brown is Director, Australian Institute for Relationship
Studies, which is affiliated with Relationships Australia (NSW),
135 New South Head Road, Edgecliff NSW 2027.

109

Brown

relationship, calling it a triadic modality. They argue

that this manner of questioning breaks a rule that they
believe operates in dysfunctional families: that family
members do not comment on the relationship of other
family members in their presence.
Another contribution of the Milan Associates, drawing
on the work of Watzlawick, is an elaboration of the idea
that all behaviour is communication. One family member is validly able to comment on the relationship
between two other family members by observing their
interaction, rather than by inferring their internal states.
They draw the distinction between the verb to be and
the verb to show; instead of speaking about being
depressed, for example, the therapist speaks about
showing depression.
The Milan Associates further suggest that it is useful
to focus on behavioural sequences and each persons
interpretation of behaviour, as a way of establishing circularity and new meaning. A further recommendation
to help create difference is that family members be
asked to rank each other on specific behaviours, making
discrepancies in the views of various family members
more noticeable. They highlight the usefulness of ascertaining the differences in behaviour before and after significant family events, and of asking hypothetically what
differences might become apparent if other circumstances existed in the family. These ideas about questioning were revolutionary, as were the Milan Associates unorthodox ways of interviewing, resulting in
many unpredicted changes in families. Their paper
sparked a number of other significant contributions
which elaborated and extended these ideas.
In her attempts to define a model for interviewing,
Penn (1982) concentrated on designing questions that
create difference. She stated:

the questioning technique that students had difficulty

grasping circular interviewing. So they developed a
taxonomy specifically to facilitate their teaching,
presenting four types of questions: 1. Problem definition, which is the beginning point for student inquiry
in the interview. 2. Sequence of interaction, the next
step in the interview once the problem definition is
obtained. This category, which together with the next
(comparison/classification) includes questions that
focus on differences and changes over time, agreement
and disagreement between family members and explanations for the views expressed, is very similar to
questions
delineated
by
Penn
(1982).
3.
Comparison/Classification questions are the next suggested focus in the interview for students. They include
comparisons and classifications of behaviour, beliefs,
values, thoughts, feelings and relationships. 4. Intervention is the final category, where students ask questions that, among other things, encourage clients to try
new behaviour that challenges family beliefs and rules.
These questions have embedded reframes to enable family members to see their situation differently. The
above categories each develop questions related to time,
from past to future. Fleuridas et al. provided a guide
for the interview with many examples of the types of
questions that could be used. Their classification was
thus an extremely useful learning tool. The first two categories of problem definition and sequence of interaction are very prescriptive, providing an excellent framework for students to follow when learning circular
questioning and conducting interviews early in the therapy. However, with so many possible questions that can
be asked under the comparison/classification category,
the level of guidance changes significantly from that provided with their first two categories.
Tomm (1984), in his elaboration of the Milan systemic
approach, focused on questions that highlight the spatial (differences between people, relationships, ideas,
values and perceptions) and temporal (differences over
time). His ideas continued to develop through various
other papers, culminating in his 1988 paper, where he
presented a model of four major types, based on the
theoretical assumptions underlying the questions (lineal
versus circular) and the intention of the therapist in asking them (influencing the client versus orienting the
therapist). The question types are: 1. Lineal (e.g. problem explanation and problem definition): useful in providing basic information regarding the problem, but can
convey a judgmental attitude, stimulating defensiveness
and guilt in clients. They may also validate pre-existing
beliefs around the problem. 2. Circular questions (eg.
behavioural effect and difference questions) are less
judgmental and have liberating effects of families. 3.
Strategic questions (e.g. leading and confrontational),
while being corrective in nature, have the effect of
imposing therapist views on families, with the possible
effect of creating shame and guilt. 4. Reflexive questions
(e.g. hypothetical future and observer perspective) are
seen as facilitative, providing indirect influence on clients in that they encourage reflection on the impli-

The aim of circular questioning is to fix the point in the

history of the system when important coalitions underwent
a shift and the subsequent adaptation of that shift became
problematic for the family. The information sought by circular questions is the differences in relationships the family
has experienced before and after the problem began.
(Penn, 1982: 272)

Thus, she presents a number of categories of circular

questions that highlight differences, including: 1. Differences between verbal and analogic information provided by the clients, 2. Differences between the problem definition and explanations of the problem over
time, 3. Differences between closeness of various family
members over time, and 4. Differences created through
the classification and comparison of family relational
patterns. These questions highlight the importance of
creating questions that mark difference and change over
time as well as difference between the views of family
members, particularly around the presenting problem.
Her emphasis on difference as an underlying concept
for all questions asked in the interview was an important
contribution to circular interviewing.
Fleuridas et al. (1986) also focussed on differences
and change over time as a way of gaining information
for the therapist. However, they noted when teaching
110

A.N.Z.J. Fam. Ther., 1997, Vol. 18, No. 2

cations of current behaviour and ideas, and a consideration of new options.

Tomms notions about therapist intent and assumptions in questioning are very useful for experienced
clinicians, who are thus encouraged to monitor the
effect of their questions during the interview, and to ask
different questions if their intentions are not realised.
Tomm also emphasises the idea of creating a balance
between questions and statements. If there are too many
questions, clients may feel interrogated, while if there
are too few, therapists may experience restrictions in
the information available. Tomms emphasis on the
statement/question balance is particularly useful when
training therapists who come from client centred therapies that emphasise reflecting client statements, as
opposed to asking questions. However, as student therapists grapple with learning circular questioning, the new
model presented by Tomm is often overwhelming.
MacKinnon (1988) focuses on the Milan questioning
process as a way of exploring the openings presented
during a session. Openings are indicated by key words
and utterances that signify emotionally laden areas. The
therapist uses questioning to explore these areas and
facilitate increased intensity and emotional openness.
Questioning is also used to explore and shift premises
that constrain family members relationships and ability
to resolve conflictual or painful issues.

conceptualise difference as a type of question (eg.

Tomm, 1988; Fleuridas et al., 1986). When the purpose
of an interview is to provide information, and information is difference, then having a category of difference questions is somewhat ambiguous, as all questions
should have the goal of providing information through
the creation of difference. As Selvini Palazzoli et al. point
out, 1. Information is a difference. 2. Difference is a
relationship (or a change in the relationship) (1980: 8).
Having a goal of creating difference in the information
provided by clients through the subsequent questions
asked in a therapy session is thus fundamental to the
interviewing style. For example, clients often provide
non-specific information such as The problem has
always been present or We all think the problem is
caused by bad genes. When a therapist operates from
the underlying principle of creating difference in the
session, he or she can watch for such comments and
then turn them into more productive information, by
asking questions such as Who most accepts the view
that the problem has always been there? or When has
the problem been better/worse?; Who most thinks that
the problem is caused by bad genes? or When did the
family begin to think that the problem was caused by
bad genes? Thus, if the therapist has an underlying goal
of creating difference and hence information, unhelpful
statements from clients can easily be turned into useful
information that uncovers subtle differences ignored by
the family, providing an alternative view to a dominant
one presented by clients in the interview.
As Tomm (1985) notes, difference can be created
categorically (between people, relationships, perceptions and events) and temporally (between past, present
and future). Differences across time, people and events
are very useful categories to use for training therapists
because there is little overlap between them, which creates less confusion than a more extensive list of specific
difference question types might provide. When training
students, I have used some of Tomms categories,
adding one of my own, and focusing on the following
categories as the basic underlying principle for interviewing in a way that maximises difference:

THE TWO FACTOR MODEL: CREATING

DIFFERENCES AND CONNECTIONS
Most of the other models of questioning focus on types
of questions as a way of teaching student therapists.
Such an approach requires students to learn a classification system and the specific questions in each category, in order to apply them when conducting therapy.
By contrast the two factor model emphasises a process
for asking questions, rather than relying on students to
memorise specific categories of questions. While both
approaches are valid, I would argue that learning how
to construct questions may be more appropriate for the
beginning therapist than memorising categories of questions. Following a period of constructing their own
questions, it may then be useful for new therapists to
reflect on the categories of questions that exist in various models. This approach encourages students to draw
questions from a range of models.

1. Across time
Difference over time highlights the changes that occur,
which clients often fail to notice, possibly due to their
subjective involvement, anxiety or beliefs about a problem. Contrasting past time with the present is a way of
creating difference. For example, when asked about
these differences over time, a client may note that a
problem began at a particular point, or that a behaviour
is more extreme during one week than another. Future
time, which is always hypothetical, can also be used as
a way of pointing to new possibilities and the creation
of hope, as is well documented elsewhere (Penn, 1985).
For example, when asked about differences in the
future, a client may think that a problem will be
resolved at some future date or that other changes will
occur that will alter the nature of the problem.

Creating Differences
When clients generalise a particular problem situation
to their entire life, it is often helpful to deconstruct the
generalisation by asking questions that create difference.
The concept of difference is crucial in all of the models
of circular questioning discussed and in this model, is
best introduced as the first factor for emphasis when
teaching circular questioning. Penns (1982) model is
an example of one that has difference as a fundamental
principle underlying all question types. Other models
111

Brown

Thinking of your head and your heart, which part

wants to stay in the relationship and which part
wants to leave?
How much of you is angry and how much of you is
sad about your grandmothers death?
Do you think he is ruled more by his feelings or by
his thoughts when he stays out late? What do you
think those thoughts are?
When you feel angry with her, what part of you taps
into your current feelings about the situation and
what part taps into feelings from the past?
When you argue with each other, which side is more
likely to take over, your rational side or your
emotional side?

Examples:

When did the problem begin?

When is the problem most difficult?
When is the problem least evident?
When might the problem be solved?
Will the problem be solved more quickly by itself or
more quickly if you have some assistance?
How long will it take for this problem to go away?
How might the problem change in the future?

2. Between people
Differences between people highlight unique behaviours, attitudes or beliefs that clients may fail to notice
when wanting to create a greater sense of unity or
cohesion. For example, clients may state that all family
members are sad about the death of a grandparent, but
fail to acknowledge, until questioned on this difference,
that one family member is having a much stronger reaction to the death. On the other hand, using questions
that rate family members on particular behaviours, attitudes or beliefs and thus creating difference, may draw
out similarities that have been ignored by clients. For
example, family members may see one person as being
argumentative until asked to rank all members on their
argumentativeness; then they realise that it is a common
behaviour in the family.
Examples:

4. Between situations
Differences between situations highlight differences in
client behaviour, attitudes and beliefs, depending on the
context. Client flexibility and control over behaviour,
attitudes or beliefs are emphasised when questioning
reveals these differences between the same persons
behaviour in different contexts.
Examples:
Is he more likely to be violent in public or private?
Why do you think that is the case?
Do you think her behaviour is better at home or at
school?
Do you think her attitude about you being the breadwinner is more evident when you are with friends
or when you are with your family?
In what situation is the problem most noticeable?
Where are your religious beliefs most obvious to
others?
Where do you think you are most yourself, at work
or at home?

Who is the most argumentative in the family? Who

next?
Who most believes that parents should make all the
rules in the family? Who next?
Who misses John the most when he is away at boarding school?
Who between the two of you most thinks that
women should have the major responsibility for
the housework?
Who is most committed to the relationship? How do
they show it?
Who most shows their concern about the problem?
If the problem stays around for a long time, who will
be most concerned about it?

These four categories provide a helpful framework for

understanding how news of difference can be created
in an interview, as a way of increasing the information
value of the interview for therapist as well as client.
Student therapists are able to practise asking questions
that highlight difference on the above dimensions in
response to client statements that normally mask any
notion of difference. The skill in creating difference is
an important task for students in the process of learning
circular questioning.

3. Between parts of a person

Differences within a person highlight the multidimensional nature of behaviour, attitudes and beliefs and the
absurdity of categorising a person with a particular
label. When a person is seen to have a particularly
strong behaviour, attitude or belief, it is useful to distinguish between the part of him or her that thinks or
behaves one way and the part that thinks or behaves
another way. This recognises the complexity of human
nature, and the potential for tapping into the ambivalence that may be there.
Examples:

Drawing Connections
When clients are overwhelmed by a problem, they often
see it as isolated from the context in which it exists.
Their focus may be narrowed by the gravity of the situation. Explanations of the problem are frequently linear,
with the problem being seen in a simplistic black and
white way. Ruesch and Bateson note that . . . all knowledge of external events is derived from the relationships between them . . . (1968: 173). Seeing the pattern that connects that Bateson writes about puts the
problem in a much broader context, changing its linear

Is your dads tender side or his gruff side easier for

you to see? Where do you think the tender side is
hidden?
112

A.N.Z.J. Fam. Ther., 1997, Vol. 18, No. 2

focus to one that is systemic. This more complete picture of client problems is helpful for the therapist and
client in creating understanding and facilitating change.
In teaching students to draw connections, I find it
useful to use content categories in the first instance. The
content categories selected have been delineated elsewhere (Brown, 1997), but will be discussed here briefly.
The categories of content selected provide a trigger for
students to stimulate questions within the broad category to use for drawing connections or links.

personal feelings. This is particularly important early on

in the therapeutic process, while later on, questions suggesting a greater control of feelings by the client may
be more therapeutic. Feeling questions can be linked to
behaviour, beliefs, meaning and relationships. Note the
difference between these questions, which imply less
control over feelings, and the more confronting how do
you feel? questions often asked by beginning therapists.
Examples:
What feelings come up for you when she calls you
incompetent?
When he tells you that he doesnt love you any more,
what emotions emerge?
When she feels sad, what feelings do you have?
How do you feel, knowing that he believes women
should stay at home and raise the children?
What sense do you make of his anger that you will
not have sex with him?
When you spend long periods of feeling angry with
each other, how do you think that affects your
relationship?

1. Behaviour
Questions focusing on behaviour are useful in getting a
good understanding of exactly what is happening with
clients. Often a vagueness in client report of problems
can be clarified through emphasising specific behaviours. Behaviours can be linked to other behaviours, to
feelings, to beliefs, to meaning and to relationships.
Questions focusing on behaviour are also useful as a way
of maintaining neutrality, as value laden categories are
frequently avoided when the focus is on behaviour. For
example Why do you spend the day in bed? is more
neutral than Why are you so lazy?.
Examples:

3. Beliefs
Questions about beliefs are asked to help understand
some of the underlying ideas that clients have which
influence the way they act, think and feel. Beliefs tend
to emerge from past experience in situations with significant people such as family, friends and former lovers.
They frequently result in clients accepting a certain way
of acting, thinking and feeling as the unquestioned
norm. Beliefs also may arise out of painful past experience and thus operate as a way of protecting the self.
Beliefs are often attributed to participation in groups
such as family of origin, culture, social class, gender,
sexual orientation and religion. Beliefs can be linked
with behaviour, feelings, meaning and relationships.
Examples:

What happened after you asked him to go to his

room? What happened then? . . . How did it all end?
When she stays out late without telling you where
she is, what feelings come up for you?
How does hitting your son relate to your beliefs
about child rearing?
How do the ways you fight relate to the ideas about
how couples resolve arguments that you brought
from your family?
How do you make sense of the difference between
his behaviour at home and at school?
When the family argue, does it bring you all closer
together or further apart?

When he is reluctant to help out with the housework, what do you think he believes about the role
of women?
When she is sad about not having had a child, do
you think it is more because she wants a child or
because her family wants a grandchild?
When you believe that conflict should be out in the
open in any relationship, does he agree more or less
with you?
How do you make sense of her saying on the one
hand that she believes men should be responsible
for their own feelings and yet on the other hand continuing to seek out your deeper emotions?
When you believe that there is a danger in expressing your anger at home, do you think that unites or
separates the family?

2. Feelings
Questions about feelings are asked as a way of highlighting the emotional experience of family members,
where there is often a tendency to focus on behaviour
and meaning. While these questions are useful in
increasing therapeutic understanding of the meaning
and motivations of behaviour, they are also a way of
increasing client self understanding, as well as stimulating increased empathy and understanding by other family members. The way that feeling questions are asked
may unintentionally imply that clients have too little or
too much control over their feelings. For example the
well-known question How does that make you feel?
implies that a client has no control over how he or she
feels. The question When did you decide to become
depressed? may imply so much control that the client
feels misunderstood. Thus, feeling questions need to
carefully weave a path between these two extremes, by
maintaining therapeutic connection with the client
while not inferring that she or he has no control over

4. Meaning
Meaning questions tap into the way that clients interpret
their world through their interaction with others. These
113

Brown

interpretations are useful in helping to understand their

behaviour in its interpersonal context. They are useful
for understanding self as well as other family members.
When they are linked to behaviour, feelings, beliefs and
relationships, questions about meaning are powerful in
creating change in family systems.
Examples:

ing principle of circular questioning is a process of

creating distinctions and connections can be a useful
simplification of the process, allowing students to
experiment. In this way, the usefulness of various questions will become apparent very quickly to students and
they will thus become self motivating. Furthermore,
linking questioning style to aspects of neutrality and
hypothesising then creates a more complete context for
conducting an interview in a circular style. As experience builds up at this level, it is possible for students
to focus on some of the more complex models of questioning and integrate these into their repertoire.

How do you understand it when he says that he

loves you and he neglects to help you when you ask
for assistance?
When he is feeling sad, what do you think his sadness is about?
When she thinks that your anger relates to unresolved issues with your mother rather than her, how
do you think that she gets that idea?
When you say that his behaviour shows he doesnt
want to live at home any more, how do you think
he sees that statement?
When she says that she feels you do not love her
any more, how do you think that affects your
relationship?

References
Brown, J., 1997. The Question Cube: a Model for Developing Question
Repertoire in Training Couple and Family Therapists, Journal of
Marital and Family Therapy, 23, 1: 2740.
Fleuridas, C., Nelson, T. and Rosenthal, D., 1986. The Evolution of
Circular Questions: Training Family Therapists, Journal of Marital
and Family Therapy, 12, 2: 113127.
MacKinnon, L., 1988. Openings: Using Questions Therapeutically,
Dulwich Centre Newsletter, Winter, 1518.
Penn, P., 1982. Circular Questioning, Family Process, 21, 3, 267280.
Penn, P., 1985. Feed-forward: Future Questions, Future Maps. Family
Process, 24, 3: 299310.
Ruesch, J. and Bateson, G., 1968. Communication: The Social Matrix
of Psychiatry, New York, Norton.
Selvini Palazzoli, M., Boscolo, L., Cecchin, G. and Prata, G., 1980.
HypothesisingCircularityNeutrality: Three Guidelines for the
Conductor of the Session, Family Process, 19, 1: 312.
Tomm, K., 1984. One Perspective on the Milan Systemic Approach:
Part II. Description of Session Format, Interviewing Style and Interventions. Journal of Marital and Family Therapy, 10, 3: 253271.
Tomm, K., 1985. Chapter Four: Circular Interviewing: A Multifaceted
Clinical Tool. In D. Campbell and R. Draper (Eds), Applications of
Systemic Family Therapy, New York, Grune and Stratton.
Tomm, K., 1987. Interventive Interviewing: Part II. Reflexive Questioning as a Means to Enable Self-healing, Family Process, 26:
167183.
Tomm, K., 1988. Interventive Interviewing: Part III. Intending to Ask
Lineal, Circular, Strategic or Reflexive Questions, Family Process,
27, 1: 115.

5. Relationship
Relationship questions are used to allow family members to comment on the nature of their relationships
with each other, as they experience them through their
every day interaction. These comments are rarely made
outside the therapy room. However, assumptions about
the nature of relationships are frequently made and
acted upon in the context of the family. Making these
assumptions explicit through relationship questions that
draw connections between relationships and behaviour,
feelings, beliefs and meaning can create significant new
understanding, and thus provide the impetus for change
in the family.
Examples:

When she says that she is coming home right after

school is finished and does not do so, how do you
think that affects the nature of your relationship?
When he is angry and expresses that to you, does
that make you feel closer together or further apart?
When you believe that relationships should not have
any conflict and conflict continues, what does that
say about your relationship?
When she is pessimistic about your willingness to be
more involved in the family, how do you think that
affects your relationship?
When he feels that the family is working together
well, are you more likely to feel closer to the family
or more distant?

I had written him a letter which I had,

for want of better
Knowledge, sent to where I met him
down the Lachlan, years ago,
He was shearing when I knew him, so I
sent the letter to him,
Just on spec, addressed as follows,
Clancy, of the Overflow
Clancy was lucky. He received his letter.
Mails were better then, perhaps. You may
not be as lucky, unless you advise
Blackwells of your address changes.

CONCLUSION

READ INSIDE FRONT COVER FOR

DETAILS.

While a number of models exist for teaching circular

questioning, they are frquently difficult to understand
for new therapists. Understanding that the major guid114