The purpose of this paper is to examine the issue of behavioral
reflexivity from a behavior-analytic perspective. Two quotations from
behavioral researchers are first considered, and both suggest that
behavioral reflexivity is an issue best ignored. The nature of
behavioral reflexivity is then examined, in detail, by dividing it into
three basic assumptions. This examination suggests that behavioral
reflexivity precludes the possibility of finding an ontological
(correspondence-based) truth in behavior analysis, and therefore the
ontological truth of behavioral reflexivity itself is undermined. If a
pragmatic truth criterion is adopted, however, the truth of behavioral
reflexivity must be defined in terms of its usefulness in achieving
particular goals, and thus the ontological truth of behavioral
reflexivity becomes irrelevant. As a starting point for demonstrating
the usefulness of behavioral reflexivity, an interpretive behavior
analysis of behavioral reflexivity is conducted. This analysis suggests
that behavioral reflexivity is produced, in large part, by the
contingencies operating in the verbal community that establish
responding to one's own behavior, and to "truth," in accordance with the
relational frames of HERE and THERE, and NOW and THEN. The analysis is
then used as the basis for an exercise for teaching students about
behavioral reflexivity. Insofar as this exercise is useful in achieving
particular goals, behavioral reflexivity is a true and welcome feature
of behavior analysis.

It would be absurd for the behaviorist to contend that he is in any way
exempt from his analysis. He cannot step outside of the causal stream
and observe behavior from some special point of vantage, 'perched on the
epicycle of Mercury.' In the very act of analyzing human behavior he is
behaving. (Skinner, 1974, p. 234)

The above quotation suggests that behavior analytic epistemology
considers the scientific activity of the behavior analyst to be part of
a behavioral stream, and thus the science of behavior analysis consists
of the very same stuff (behavior) that it takes as its subject matter.
This position, which we shall call behavioral reflexivity,[1] has
created considerable discomfort within the behavior analytic community
over the years. Schnaitter (1987), for example, has argued that
behavioral reflexivity, though obviously correct, is a conceptual issue
best avoided:

Doing science is performance under constraint, subject to its own
analysis consisting of further performance under constraint, and so on.
Just as one cannot perceive outside one's perceptions or think outside
one's thoughts, so one cannot behave or perform outside one's behavior
or performance, even when doing science. Be that as it may, more than a
few well-intentioned behaviorists have been trapped by this captivating
eddy in the behavioral stream. Undue engagement with this truism can
endanger any sort of productive analysis of behavior. (pp. 5-6)

Consider also the following passage written by Staddon (1993) in which
he argued that behavioral reflexivity should be removed altogether from
behavior analytic epistemology:

[Others] may object that we, the experimenters, are also organisms,
subject to the vaunted "laws" of behavior. We must (they contend)
consider the implications of our laws for our own behavior. This is a
surprisingly persuasive argument for many behavior analysts, but I
believe it also is mistaken. ... Subject (the scientist) and object (the
thing he or she is studying) are different, and we have no evidence that
a strategy that blurs the difference will advance any science. (p. 247)

Despite the concern expressed by the foregoing authors with regard to
behavioral reflexivity, we are of the firm opinion that the best way to
deal with behavioral reflexivity is to embrace it fully, not as a
conceptual, philosophical, or logical position, but as an example of the
behavior of the scientist. In so doing, we hope to show that behavioral
reflexivity is an important part of the "language game" we call behavior
analysis.

In the first part of this paper we outline what we consider to be the
nature of behavioral reflexivity by dividing it into three basic
assumptions. Having outlined these assumptions, we then consider the
implications of pragmatism for behavioral reflexivity, and we argue that
the truth of behavioral reflexivity must be defined in terms of its
usefulness in achieving particular goals. To demonstrate the usefulness
of behavioral reflexivity, we conduct an interpretive behavior analysis
of behavioral reflexivity that facilitates, in our opinion, a behavior
analytic understanding of certain types of scientific behavior. Finally,
we outline an exercise that we have used for teaching students about
behavioral reflexivity.

The Three Assumptions of Behavioral Reflexivity <#toc>


        Assumption 1: What is Known is Always a Behavioral Function

In behavior analysis, all events are known or defined in terms of
behavioral function, rather than as physical things that exist
independently of behavior (see Skinner, 1938, on the inseparability of
stimuli and responses; see also Barnes & Roche, 1994; Hayes, L. J.,
1993). To understand this assumption of behavioral reflexivity, consider
the simple case of a wedged-shaped piece of metal. In common-sense
terms, the wedge is a physical thing that exists independently of
behavior. In behavior analysis, however, the wedge is defined only in
terms of its behavioral functions that emerge in a particular stream of
behavioral interactions (see Barnes, 1989, pp. 340-341; Schoenfeld &
Farmer, 1970). The wedge, for instance, may be defined as a
discriminative stimulus for a particular response, such as picking it
up, or uttering "that's a metal wedge," or alternatively it may defined
as a reinforcing stimulus for other responses, such as pointing at the
wedge, or uttering "pass me the wedge." Even if the wedge in question is
considered to be "imaginary" (e.g., a wedge with magical properties in a
children's story book), such a wedge would still be defined
behaviorally; in this case, however, terms such as autoclitic or
metaphorical extension (Skinner, 1957) might be used to define the
wedge. In behavior analysis, therefore, the wedge (or any other part of
the universe) is always defined or known within a particular behavioral
stream (see Barnes & Roche, 1994).


        Assumption 2: The Activity of Each Organism Participates in a
        Separate Behavioral Stream

Behavioral reflexivity does not allow one behavioral stream to overlap
with, or intrude upon, a second behavioral stream. Consider, for
example, a driver who stops at a red traffic light. In behavior
analysis, we might define the red light as a discriminative stimulus for
pressing on the brake, and we would explain this discriminative function
in terms of the history of behavioral interactions that established the
function. This history of behavioral interactions and the functional
relation thus obtained between the red light and the braking response
constitutes part of the driver's behavioral stream. Consider now, a
second driver who also stops at the red traffic light. In common-sense
terms, we would likely say that the red light is the same red light for
both drivers. In behavior analysis, however, we must define two separate
discriminative functions[2] for the two braking responses emitted by the
drivers. In a sense, there are two red lights--one in each behavioral
stream. In summary, therefore, Assumption 2 of behavioral reflexivity
views the activity of each organism as participating in separate
behavioral streams. Even when two organisms are responding to the "same"
event (in common-sense terms), they do so from within their respective
streams.


        Assumption 3: The Scientific Activity of The Behavioral Analyst
        Participates in a Behavioral Stream

According to the quotation presented at the beginning of this paper,
even the scientific activity of a behavior analyst participates in a
behavioral stream and thus he or she cannot "observe behavior from some
special point of vantage, 'perched on the epicycle of Mercury.'"
(Skinner, 1974, p. 234). From this perspective, scientific activity does
not involve discovering the fundamental laws of nature or developing an
increasingly accurate picture of an ontological reality; instead,
scientific activity itself is subject to a behavior analysis. Again, in
Skinner's (1969) words:

Scientific laws ... specify or imply responses and consequences. They
are not ... obeyed by nature but by men that deal effectively with
nature. The formula s = 1/2 gt^2 does not govern the behavior of falling
bodies, it governs those who correctly predict the position of falling
bodies at given times. (p. 141)

In behavior analysis, therefore, the output from a cumulative record,
for example, is not a representation of what the subject "really" did in
the operant chamber, but is rather a discriminative stimulus for a
particular "scientific" response, such as "scallop" or "break-and-run,"
that has been differentially reinforced in the presence of that pattern.
In effect, the scientific activity of a particular behavior analyst is
always part of that scientist's behavioral stream.

Truth and Behavioral Reflexivity <#toc>

Assumptions 1 and 2 of behavioral reflexivity, though strange from a
common-sense point of view, create little difficulty for behavior
analysis. Assumption 3 of behavioral reflexivity, however, appears to
preclude the possibility, in behavior analysis, of finding a truth that
corresponds to an ontological reality. Furthermore, and more importantly
for the current paper, Assumption 3 appears to undermine behavioral
reflexivity itself. Consider the following. If the scientific activity
of the behavior analyst is the product of that analyst's behavioral
history, then he or she can never claim to have found an ontological
truth, because a different or more extended history may have produced a
different truth (an ontological truth, by definition, is immutable,
absolute, and final). Furthermore, the argument that behavior analysis
is reflexive cannot be true in an ontological sense because the argument
is simply the product of a particular behavioral history. Behavioral
reflexivity thus appears to sow the seeds of its own destruction.
Nevertheless, truth in behavior analysis, including the truth of
behavioral reflexivity itself, may be rescued if a pragmatic truth
criterion is adopted. The rest of the current section is devoted to this
issue.

When confronted with the problem of truth in behavior analysis, the most
typical response is to point out that we never attempt to construct a
scientific analysis that corresponds to an extant, ontological reality.
Instead, behavior analysis is concerned only with a pragmatic version of
the truth. From this perspective, truth is defined simply in terms of
prediction and control (i.e., successful working). If a scientific
statement is useful in helping the behavioral scientist to achieve the
goals of prediction and control, with some degree of scope and
precision, then the statement is considered true (see Barnes & Roche,
1994; Hayes, S.C., & Brownstein, 1986; Skinner, 1974, p. 235). The
correspondence between the scientific statement and an ontological
reality is completely irrelevant.[3]

In appealing to the pragmatic truth criterion of successful working, it
appears that behavior analysis circumvents the problem of never finding
an ontological truth. In other words, if behavior analysis is
unconcerned with the ontological truth, then it can hardly be criticized
for being unable to find such a truth. A point sometimes missed,
however, is that although pragmatism addresses the issue of never
finding an ontological truth in behavior analysis, the truth criterion
of successful working does not undermine behavioral reflexivity (see
Hayes, L. J., 1993). In other words, the scientific goals of prediction
and control do not allow the behavior analyst to step outside the
behavioral stream, and make contact with a nonbehavioral, ontological
reality. Consider the following two points.

First, stating that one's scientific goals are prediction and control,
and then trying to achieve them must also be viewed in behavior analysis
as behavioral events. Scientific goals cannot be considered objective
truths handed down on tablets of stone from the god of ontology. In
behavior analysis, therefore, the goals of a particular scientist are
produced by that scientist's history of behavioral interactions, and as
such the goals are part of a unique behavioral stream. In the words of
Leigland (1993):

How is it a particular scientist has actually chosen a particular set of
goals over another? From the perspective of radical behaviorism, the
answer ... is a matter of the variables, experiences, and contingencies
that constitute an individual's history of contact with the social and
non-social environment, and is not to be confused with the verbal
reasons which we would normally provide as "explanations" (whether in
the context of casual or philosophical discourse) for what we choose.
(p. 30)

In a sense, therefore, the scientific goals of prediction and control
are purely an accident of history. If, for example, a particular
behavior analyst had been taught by different professors who emphasized
prediction over control, the analyst may have become a different type of
psychologist or perhaps pursued an entirely different career.
Furthermore, even if two specific behavior analysts claim and pursue the
goals of prediction and control, their different histories may lead each
of them to "interpret" these goals differently (see Hayes, L. J., 1993,
p. 4041). Consider, for example, the different approaches suggested by
Murray Sidman and Steven Hayes to the issue of stimulus equivalence (see
Barnes, 1994; Barnes & Roche, 1996).

Second, in adopting the pragmatic goals of prediction and control a
behavior analyst may often talk as if "real" events were being contacted
outside of the behavioral stream, but according to behavioral
reflexivity such talk is ultimately considered to be a particular
instance of scientific verbal behavior. Imagine, for the sake of
argument, that in the course of an experimental analysis a behavior
analyst discovers that whenever he or she arranges for event X to occur,
event Y always follows. If the analyst then states that "X produces goal
Y," surely this requires a correspondence between the goal-statement and
the X-then-Y event? According to behavioral reflexivity, however, the
goal-statement is a behavioral event, and thus no correspondence between
the statement and a nonbehavioral, ontological reality need be assumed.
In technical terms, we might interpret the goal-statement as follows: A
history of arbitrarily applicable relational responding established a
transformation of functions in accordance with the mutually entailed
relation of coordination between the event, X followed by Y, and the
goal-statement "X produces goal Y" This technical vocabulary highlights
that the goal statement is treated as a behavioral event that is
functionally related to other behavioral events in a behavioral stream.
The issue of correspondence, therefore, is simply irrelevant. Of course,
even the foregoing technical interpretation itself is subject to further
analysis, and thus it too does not necessarily correspond to an
ontological reality.[4] (The requirement that analytic statements in
behavior analysis are always subject to further analysis does, however,
create a problem for behavioral reflexivity, and we shall address this
issue in the final paragraph of this section).

Having seen that pragmatism does not undermine behavioral reflexivity,
we now come to the truth of behavioral reflexivity itself. As suggested
at the beginning of this section, the argument for behavioral
reflexivity cannot be true ontologically. Instead, the argument for
behavioral reflexivity is a verbal event that is functionally related to
other verbal and nonverbal behavioral events that go to make up a
particular behavioral stream, and as such the truth of the argument
rests solely on its functional utility as defined in that stream. The
question is therefore; what goal or goals have we, the authors, achieved
in focusing on the issue of behavioral reflexivity? We have two answers
to this question. First, having conducted an interpretive behavior
analysis of behavioral reflexivity itself, presented in the next
section, we now understand, more clearly than hitherto, certain
properties of scientific behavior. Second, we have developed an exercise
for teaching students about behavioral reflexivity that appears to
increase their approval of, and interest in, behavior analysis.

At this point we might be criticized for giving the impression that, in
analyzing and/or teaching behavioral reflexivity, one somehow steps
outside of one's behavioral stream and makes direct contact with a
"real" ontological event called behavioral reflexivity. This, however,
was not our intention. According to behavioral reflexivity, when
analyzing or teaching one is behaving--even in reading this sentence
you, the reader, are behaving. The putative ubiquity of one's own
behavior brings us to the most troublesome feature of the truth of
behavioral reflexivity. The trouble appears to be that the ubiquity of
one's own behavior cannot be analyzed, at least in the traditional
sense. To determine that one is always behaving requires that one
behaves, and then analyzes that behavior without engaging in any further
behavior! Trying to achieve this goal leads to a form analytic tail
chasing. Having analyzed one's own behavior, one cannot conclude that
one is always behaving because the act of analyzing one's own behavior
remains unanalyzed. One must, therefore, analyze one's analysis, and
then one is forced to analyze that analysis of the analysis, and so on,
ad infinitum. The idea that one is always behaving may therefore lead to
a never ending cycle of futile analyses, and as such the utility or
truth of behavioral reflexivity is apparently threatened. in the next
two sections, however, we will attempt to show that the ubiquity of
behavior suggested by behavioral reflexivity may, in fact, serve a
useful purpose.

A Behavior Analytic Interpretation of Behavioral Reflexivity <#toc>

We will now conduct an interpretive behavior analysis of behavioral
reflexivity itself, focusing primarily on Assumption 3 and the ubiquity
of behavior. We focus on these issues because they have attracted the
most attention over the years, and in a sense they go to the very heart
of behavioral reflexivity. The analysis will draw heavily on relational
frame theory, and the ideas expressed in the paper Making Sense of
Spirituality by Steven Hayes (1984). In essence, we will argue that the
contingencies that establish responding to one's own behavior in
accordance with the relational frames of "HERE and THERE" and "NOW and
THEN" are also involved in producing behavioral reflexivity, and in
particular the tail-chasing phenomenon.


        Relational Frame Theory and the Frames of HERE and THERE, and
        NOW and THEN

According to relational frame theory,[5] learning to name objects and
events in the world represents one of the earliest and most important
forms of arbitrarily applicable relational responding. For instance,
parents often utter the name of an object in the presence of their young
child and then reinforce any orienting response that occurs towards the
named object. This interaction may be described as, hear name A look at
object B. Parents also often present an object to their young child and
then model and reinforce an appropriate "tact" (Skinner, 1957). This
interaction may be described as see object B right arrow hear and say
name A (see Barnes, 1994, for a detailed discussion). Initially, each
interaction may require explicit reinforcement for it to become firmly
established in the behavioral repertoire of the child, but after a
number of exemplars have been trained, derived "naming" may be possible.
Suppose, for example, a child with this naming history is told "This is
your shoe." Contextual cues, such as the word "is" and the naming
context more generally, may establish a transformation of pointing and
speaking functions, for the shoe, in accordance with the mutually
entailed relation of coordination. Without further training, for
example, the child will now point to the shoe when asked "Where is your
shoe?" (name A right arrow object B) and will utter "shoe" when
presented with the shoe and asked "What is this?" (object B right arrow
name A).

In addition to naming, many other patterns of arbitrarily applicable
relational responding are established during a child's interaction with
the verbal community (see Barnes, 1994; Dymond & Barnes, 1995; Hayes,
S.C., 1991; Steele & Hayes, S.C., 1991). In our opinion, however, the
most important relational frames for the current interpretation of
behavioral reflexivity are, what we will call, the relational frames of
HERE and THERE, and NOW and THEN.[6] This type of relational framing
activity is established when the verbal community teaches a young girl,
for example, to discriminate her own perspective as separate from that
of others. According to Hayes, S.C. (1984), this relational activity
emerges when the girl is taught not only to discriminate her own
behavior (i.e., to see that she sees), but to discriminate her own
behavior from a consistent locus or perspective (i.e., the child sees
that she always sees from her own perspective):

First, words such as "here" and "there" are acquired which do not refer
to a specific thing but to a relation to the child's point of view. For
example, "there" is always anywhere else but "here" and "here" is always
"from this locus or point of view". Second, children are taught to
distinguish their perspective from that of others. Young children have a
hard time with the issue of perspective. For example, young children
seated across from a doll will, when asked, report that the doll sees
what they are seeing. Gradually, however a sense of perspective emerges.
A child learns what he or she sees is seen from a perspective.
Similarly, a young child, asked what she had for breakfast, may respond
with what her brother actually ate, but an older child will not make
such a mistake. Through correction, ("No, that is what your brother ate.
What did you eat?") a child must learn to see seeing from a consistent
locus. ... Suppose a child can give correct answers to the question
"what did you x ?" where "x" is a wide variety of events such as eat,
feel, watch, and so on. The events constantly change. In our terms, the
seeing and the seeing seeing change. Only the locus does not. Thus, one
consistency between the word "you" in such questions and behavior is not
seeing or seeing seeing but the behavior of seeing that you see from a
particular locus or perspective. Thus, in some real sense, "you" are the
perspective. (Hayes, S.C., 1984, pp. 102-103; emphasis in original)

In summary, the child's interaction with the verbal community
establishes the relational frames of HERE and THERE, and NOW and THEN
and the child is taught to discriminate her own behavior as always
located in the same positions in the two relational frames (i.e., she is
always behaving HERE and NOW, and not THERE and THEN). This type of
relational framing thus generates a perspective, or context, from which
all things and events are experienced (see Figure 1).

Once the relational frames of HERE and THERE, and NOW and THEN are
established in the behavior of a particular person they become an
inherent property of most verbal events for that person. Whenever the
person talks to someone else it will be from the perspective of HERE and
NOW about events that occurred THERE and THEN. The simple greeting, "How
are you?" for example, locates the speaker HERE and NOW, and the
listener THERE and THEN. In effect, the speaker is asking, HERE and NOW,
for a response from the listener that will occur THERE (two or three
feet away) and THEN (in a second or two). Even when talking to oneself a
person responds in accordance with these relational frames. Imagine, for
example, that a speaker utters the self-deprecating statement "I'm
really stupid," having spent two hours solving what the speaker
considers to be a relatively simple problem. In this instance, the
statement locates the speaker's current behavior HERE and NOW (having
solved the problem) talking about the speaker' behavior THERE and THEN
(before the problem was solved). In summary, the relational frames of
HERE and THERE, and NOW and THEN establish a constant division between
the speaker and the spoken about. The speaker is always HERE and NOW,
and the spoken about is always THERE and THEN.


        A Relational Frame Interpretation of Behavioral Reflexivity

How may we use the concept of relational frames to construct a
behavior-analytic interpretation of behavioral reflexivity? Basically,
our argument is as follows. When a behavior analyst is confronted with
the reflexive nature of behavior analysis, the behavioral functions of
the word "truth" (as established by the wider verbal community)
repeatedly shift from HERE and NOW to THERE and THEN in the relational
frames. This shifting of truth functions may cause the behavior analyst
to deny or avoid behavioral reflexivity, or alternatively it may change
the way in which the behavior analyst responds to "truth." To appreciate
fully this interpretation, consider the following scenarios.

* Scenario 1: Relational responding to ontological truth is restricted
to one statement. * Imagine a behavior analyst who reads Skinner's
(1974) statement:

It would be absurd for the behaviorist to contend that he is in any way
exempt from his analysis. He cannot step outside of the causal stream
and observe behavior from some special point of vantage, 'perched on the
epicycle of Mercury.' In the very act of analyzing human behavior he is
behaving. (p. 234)

Having read this passage, the behavior analyst may conclude that finding
the truth in behavior analysis is impossible. This conclusion emerges
for the following reason. Truth in the behavior analyst's wider verbal
community normally depends upon the putative correspondence between
statements made HERE and NOW about an ontological reality located THERE
and THEN (see Hayes, L. J., 1993). In common sense terms, reality is
always waiting, THERE and THEN, to be described truthfully, HERE and
NOW. (Of course, untrue statements are also located HERE and NOW about
THERE and THEN, but in this case no correspondence to "reality" is
assumed). Given this verbal context, the behavior analyst therefore
starts by responding to scientific truth statements, located HERE and
NOW, as corresponding to an ontological reality, located THERE and THEN;
we will refer to this "correspondence-based" truth as ontological (see
Figure 2, upper panel). Having read Skinner's (1974) "reflexive"
passage, however, ontological truth statements, as examples of
scientific behavior, shift to THERE and THEN in the relational frames
(i.e., ontological truth is behavior waiting, THERE and THEN, to be
analyzed). Consequently, the behavior analyst is unable to utter an
ontologically true statement (HERE and Now) about a behavioral reality
(THERE and THEN) because the relational functions of HERE and NOW, that
would normally be attached to that truth statement, have been replaced
by the relational functions of THERE and THEN (Figure 2, middle panel).
In other words, the behavior analyst concludes that no scientific
statement will ever truly correspond or refer to an ontological reality
because any such statement itself will always be part of that reality.
Paradoxically, however, even in drawing this conclusion the behavior
analyst is still responding to an ontological truth as located HERE and
NOW in the relational frames. In other words, when the behavior analyst
says, "Scientific statements cannot correspond to reality," he or she
responds to this statement (located HERE and Now) as referring to an
ontological lack of correspondence between scientific statements and
reality (located THERE and THEN) (Figure 2, lower panel). At this point,
therefore, responding to ontological truth has been restricted to this
one and only statement.

* Scenario 2: The behavior analyst fails to adopt the pragmatic truth
criterion. * Having reached the point outlined in Scenario 1, imagine
that the behavior analyst now reads the following passage from Skinner
(1974):

[Scientific knowledge] is a corpus of rules for effective action, and
there is a special sense in which it could be 'true' if it yields the
most effective action possible ... (A) proposition is 'true' to the
extent that with its help the listener responds effectively to the
situation it describes. (p. 235)

At this point, the relational functions of THERE and THEN that were
attached to "reality" at the beginning of Scenario 1, are now attached
to the term "effective action." In effect, the behavior analyst
concludes that any scientific statement, located HERE and NOW, is true
to the extent that it leads to effective action (located THERE and THEN)
(Figure 3, upper panel). Note, however, that although the behavior
analyst has apparently adopted the pragmatic truth criterion of
successful working, paradoxically, he or she is responding to statements
made about effective action as ontologically true (i.e., a statement
made about effective action is perceived to be true in some absolute or
nonbehavioral sense). Consider what happens, therefore, if the behavior
analyst then reads the passage from Leigland (1993):

How is it a particular scientist has actually chosen a particular set of
goals over another? From the perspective of radical behaviorism, the
answer ... is a matter of the variables, experiences, and contingencies
that constitute an individual's history of contact with the social and
non-social environment. (p. 30)

Leigland's words lead the behavior analyst to conclude that his or her
scientific goal of effective action is not the best or most scientific
goal in some ontological sense, because a different behavioral history
may have produced a different goal, or at least a different
interpretation of that goal. In effect, the goal is not separate or
independent from the behavior of the behavior analyst. Insofar as the
truth is defined in terms of effective action, which in turn is the
product of the behavior analyst's behavioral history, the truth is again
behavior waiting, THERE and THEN, to be analyzed. In effect, the
relational functions of HERE and NOW attached to the word "truth" (in
behavior analysis) have again been replaced by the relational functions
of THERE and THEN. At this point, therefore, the behavior analyst is
forced to state that "scientific truth is a property of my own behavior"
(Hayes, S.C., 1993). In effect, relational responding to ontological
truth has been restricted to this single statement, located HERE and
NOW, about truth and behavior, located THERE and THEN (Figure 3, lower
panel). Note again, however, that although the behavior analyst has
apparently adopted the truth criterion of successful working,
paradoxically, he or she is responding to statements made about this
truth criterion itself as ontologically true (e.g., the statement,
"scientific truth is a property of my own behavior" is perceived to be
true in a literal or nonbehavioral sense). In effect, the behavior
analyst is not behaving in accordance with the truth criterion of
successful working.

* Scenario 3: Verbal tail chasing: The behavior analyst rejects either
behavioral reflexivity or ontological truth. * Having reached the point
outlined in Scenario 2, the behavior analyst's last statement may itself
be discriminated as a behavioral event (e.g., the behavior analyst may
say, "Even the statement, 'scientific truth is simply a property of my
own behavior' is more behavior."). At this point, the behavior analyst's
original statement about truth and behavior is located THERE and THEN in
the relational frames, and thus responding to the phrase "Truth is a
property of my own behavior" is no longer discriminated as ontologically
true (Figure 4, upper panel). The ontologically true statement is now
the second one uttered by the behavior analyst (i.e., "Even the
statement, 'scientific truth is simply a property of may own behavior'
is more behavior."). In effect, the behavior analyst is responding to
the second statement as located HERE and NOW, and to the first statement
as located THERE and THEN. The second statement, of course, may be
discriminated as yet more behavior, if the behavior analyst says, "But
even that statement is more behavior." (i.e., the relational functions
of HERE and NOW attached to the second statement are replaced by the
functions of THERE and THEN) (Figure 4, lower panel). This cycle of
relational responding to "ontological truth" in accordance with HERE and
THERE, and NOW and THEN may continue indefinitely, as the behavior
analyst repeatedly tries to utter an ontological truth located HERE and
NOW about his or her ontologically real behavior located THERE and THEN.
In effect, the behavior analyst chases the ever changing verbal event,
stating "but that is more behavior, and that is more behavior, and that
is more behavior ..."

This verbal tail chasing may terminate in two ways. First, the verbal
tail chase terminates when the behavior analyst states (HERE and Now)
that behavioral reflexivity (THERE and THEN) is "unproductive,"
"nonfunctional," "cannot be understood," and so forth. More
specifically, for example, the behavior analyst may state: "Although
behavioral reflexivity is true, it emerges because it is physically
impossible to analyze an event, and analyze that analysis
simultaneously, and thus we should avoid getting caught up in this
futile cycle of analyses." In effect, the relational functions of HERE
and NOW are permanently attached to an "ontologically true" statement
that defines behavioral reflexivity as unproductive, etc. (Figure 5,
upper panel). In fact, the denial or avoidance of behavioral
reflexibility is the most likely outcome, insofar as the behavior
analyst has been exposed to a long history of reinforcement for making
ontological truth statements, HERE and NOW, about events, THERE and THEN

The second way in which the verbal tail chasing may terminate is far
less likely than the first. This outcome occurs when the behavior
analyst's repeated failure to locate the ontological truth HERE and NOW
in the relational frames causes this form of relational responding to
habituate. As outlined previously, whenever the behavior analyst
discriminates an ontological truth statement as occurring HERE and NOW,
he or she subsequently discriminates the statement as another behavioral
event (THERE and THEN), and thus the statement ceases to have any
ontological truth functions. If this cycle continues, the rapid and
repeated shifting of truth statements from HERE and NOW to THERE and
THEN eventually causes this relational response to habituate, such that
any "truth" statement uttered by the behavior analyst is immediately
discriminated as a behavioral event occurring THERE and THEN. At this
point, therefore, anything the behavior analyst says or thinks is
neither true nor false ontologically. This is a strange experience for
the behavior analyst, not unlike the behavior produced by some forms of
meditation. In the same way that dispassionate observation of
spontaneous thoughts and feelings is encouraged in Buddhist forms of
meditation, for example (see Hayes, S.C., 1984), a similar effect occurs
as the behavior analyst responds HERE and NOW to each "truth" statement
as simply another behavioral event (THERE and THEN) (Figure 5, lower
panel).

Of course, the "meditative" behavior terminates sooner or later when the
behavior analyst discriminates a statement or thought as occurring HERE
and NOW, rather than THERE and THEN. At this point, however, the loss of
ontological truth functions during the meditative event may generalize
to statements made HERE and NOW, such that they are no longer
discriminated as ontologically true. In other words, responding to
multiple "truth" statements as behavioral events convinces the behavior
analyst that the search for ontological truth is a futile endeavor. For
example, the behavior analyst might state: "Having just experienced the
behavioral nature of "truth" itself, I am forced to conclude that no one
will ever know if any "truth" genuinely reflects or corresponds to an
ontological reality" As described below, it is at this point that the
behavior analyst may begin responding in accordance with the pragmatic
truth criterion, and thereby avoid any further tail-chasing behavior.

We suggested previously that when the behavior analyst apparently
adopted the pragmatic truth criterion (Scenario 2), he or she responded
to a truth statement made about "effective action" (or to a statement
made about the pragmatic truth criterion itself) as ontologically true
(i.e., the statement, located HERE and NOW, corresponded to an event,
located THERE and THEN). When such a statement was discriminated as a
behavioral event (located THERE and THEN) the statement ceased to be
true, and thus the tail chase continued as the behavior analyst tried
once more to utter a statement located HERE and NOW that corresponded to
an event located THERE and THEN. Following the meditative event and the
general loss of ontological truth functions, however, a true statement
is no longer discriminative for correspondence (between HERE and NOW,
and THERE and THEN). At this point, therefore, achieving "effective
action" becomes the only discriminative function available for
responding to a scientific statement as true (Figure 6, upper panel). In
effect, the behavior analyst now responds to a truth statement as
located HERE and NOW, in relation to an event located THERE and THEN,
because doing so is discriminated as useful in achieving a particular
scientific goal.[7] Furthermore, if such a pragmatic truth statement is
subsequently discriminated as a behavioral event (THERE and THEN), the
statement remains true because its usefulness (not its correspondence to
an ontological reality) is discriminative for calling it true (Figure 6,
lower panel). In short, by responding to true statements pragmatically,
rather than ontologically, the behavior analyst thus avoids further tail
chasing behavior-- if a statement remains true when it shifts from HERE
and NOW to THERE and THEN in the relational frames, there is no
immediate motivation for uttering another "true" statement located HERE
and NOW.

Of course, once the behavior analyst has concluded that a pragmatic
truth statement remains true after discriminating it as a behavioral
event, this conclusion itself may be discriminated as a behavioral event
(Figure 7, upper panel). Whether or not the behavior analyst engages in
any further tail chasing, however, depends upon whether he or she
discriminates such tail-chasing behavior as likely to produce effective
action in the present context (Figure 7, lower panel). In other words,
even the truth of the pragmatic truth criterion is responded to
pragmatically (i.e., in terms of effective action).


        Implications Arising From the Interpretation of Behavioral
        Reflexivity

At this point we have two outcomes arising from the pursuit of
behavioral reflexivity. The first outcome leads to the denial or
avoidance of behavioral reflexivity, whereas the latter leads to the
rejection of ontological truth. On the face of it, both outcomes produce
a behavior analyst who engages in productive scientific behavior, and
thus either option may appear equally acceptable. Some might argue,
however, that the foregoing interpretation shows that rejecting or
avoiding behavioral reflexivity is the simplest or least confusing
option[8] (see Schnaitter, 1987; Staddon, 1993; Zuriff, 1995, pp.
399-340). When teaching, for example, simply telling our students that
behavioral reflexivity is a useless dead end avoids any "futile" verbal
tail chasing and allows us to get on with the business of teaching them
about the principles of behavior analysis. If, however, we delve into
the verbal whirlpool of behavioral reflexivity, we run the risk of
confusing our students (and perhaps ourselves) and waste valuable time
on a behavioral phenomenon that is of little practical use. In contrast
to this argument, however, we have found that our pursuit and current
interpretation of behavioral reflexivity, including any resultant
confusion, may be usefully harnessed, in certain contexts, for the
purposes of teaching students about behavior analysis. The remainder of
the current paper is devoted to this issue.

An Exercise for Teaching About Behavioral Reflexivity[9] <#toc>

Students often come to psychology interested in issues that are
traditionally considered to be outside the purview of behavior analysis.
Terms such as "Self," "Meditation," "Altered States of Consciousness,"
and the like are the main "attractors." Unfortunately, however, the
consensus of opinion appears to be that behavior analysis has little to
offer students interested in these areas. Of course, we might well
bemoan the misrepresentation of our discipline by nonbehavioral
psychologists (e.g., Spinelli, 1989, p. 175) and point to various
examples of behavior analytic interpretations of "Self" and so forth
(e.g., Skinner, 1974), but is this really enough to turn the tide of
misinformation? In our experience it is not. When we first started
teaching we had students review and discuss Skinner's text "About
Behaviorism" in our tutorials, and more often than not they still
concluded that "behaviorism" was, for example, "cold," "boring," and
"simplistic." We therefore looked at what our other "more popular"
colleagues were doing in their tutorials and found that they often
engaged the students in "experiential" exercises to demonstrate the
relevance of a particular area in psychology to the students' own lives
(see Keenan, in press). More recently, therefore, we have adopted this
strategy for our tutorial work in behavior analysis.

When teaching students about behavior analytic epistemology, one
inevitably has to deal with the issue of behavioral reflexivity. In
dealing with this issue in our respective courses, we normally present a
simplified version of the material covered thus far in the current
paper, and then invite our respective students to engage in a brief
relational frame exercise. The students are informed that the exercise
is designed to help them experience their own responding to ontological
truth in accordance with the relational frames of HERE and THERE, and
NOW and THEN. The students are also told that designing and doing the
exercise with them is, in a sense, a behavior analytic response to the
"problem" of behavioral reflexivity (we will return to this issue in the
final section of the paper). Before starting the exercise, the students
are reminded that this form of relational responding is a behavior that
was established in early childhood, and it has been reinforced
incessantly by the verbal community, and as such, the behavior is
particularly difficult to extinguish.


        The Exercise: Contacting One's Own Responding to Ontological
        Truth in Accordance with the Relational Frames of HERE and
        THERE, and NOW and THEN

Teacher: I'd like us to do an exercise to show how difficult it is to
cease responding to "ontological truth" in accordance with the
relational frames of HERE and THERE, and NOW and THEN. In this way, you
should get a feeling for what our "imaginary" behavior analyst
experienced during the verbal tail chase that I just described. All I'm
going to ask you to do is to think whatever thoughts you think and to
allow them to flow, one thought after another. The purpose of the
exercise is to notice when there's a shift from locating your thoughts
THERE and THEN to locating them HERE and NOW. When your thoughts are
located THERE and THEN, in a sense, you will perceive them to be just
thoughts and as such neither true nor false, ontologically. But when
your thoughts are located HERE and NOW, in a sense, you will respond to
them as ontologically true. This sounds a bit weird, I know, but just
give the exercise a go and see what happens. I'm going to ask you to
imagine that you are sitting comfortably by the side of a stream on a
warm summers day. As you sit by the stream leaves float by on the water,
and each thought you have is a sentence written on one of these leaves.
Some people have a hard time putting thoughts into words, and they see
thoughts as images. If that applies to you, put each image on a leaf
floating down the stream. Does everyone feel happy with this so far?

Students: General agreement [if anyone seems unwilling to participate,
they may be invited to leave].

Teacher: Okay. In a minute I am going to help you relax, and then I want
you to begin letting your thoughts go by written on the floating leaves.
Now here is the task. The task is simply to watch the leaves go by
without having it stop and without finding yourself carried off down the
stream on a leaf. You are just supposed to let it flow. It is very
unlikely, however, that you will be able to do this without
interruption. And this is the key part of this exercise. At some point
you will have the sense that the stream has disappeared, or that you
have lost the point of the exercise, or that you are down in the stream,
or on a leaf, instead of being on the bank. When that happens, I would
like you to back up a few seconds and see if you can catch what you were
doing right before the stream disappeared. Then go ahead and put your
thoughts on the leaves again, until the stream disappears a second time,
and so on. The main thing is to notice when it disappears for any reason
and see if you can catch what happened right before it disappeared. Okay?

Students: General agreement [again, if anyone seems unwilling to
participate, they may be invited to leave].

Teacher: One more thing. If the stream and the leaves never get going at
all and you start thinking "it's not working", or "I'm not doing this
right" then let that thought be written on a leaf and send it down the
stream. Okay. Now let's get comfortable, close your eyes, get settled
into your chair and follow my voice. If you find yourself wandering,
just gently come back to the sound of my voice. For a moment now, turn
your attention to yourself in this room. Picture the room. Picture
yourself in this room and exactly where you are. Now begin to go inside
your skin, and get in touch with your body. Notice how you are sitting
in the chair. See if you can notice exactly the shape that is made by
the parts of your skin that touch the chair. Notice any bodily
sensations that are there. As you see each one, just sort of acknowledge
that feeling and then move on. [pause]. Now picture yourself by the side
of the stream. You stay up on the bank looking at the leaves float by.
If the stream disappears, or the leaves stop floating by, or you find
yourself on a leaf, or in the stream itself, note that, and see if you
can notice what you were doing right before that happened. Then get back
up on the bank, and let the leaves begin to float by again. Okay, let's
begin ... Whatever you think, just put it on the leaves. ... [for about
three to five minutes, allow the group to work. Don't underdo it
timewise, and use very few words. Try to read where the group are and
add a few comments as needed, like "just let it flow and notice when it
stops". Don't engage in conversation with any member of the group. If a
student opens his or her eyes calmly ask that they be closed and the
exercise be continued. If a student starts to talk, gently suggest that
even that thought be put on a leaf, saying something like, "we will talk
more about this when the exercise is finished, but for now there is no
need to talk with me. Whatever you think you want to say, let that
thought be written down and let it float by too"]. Okay, now we will let
the last few leaves float by, and we will begin to think about coming
back to this room [pause]. Now again picture yourself in this room. And
now picture the room. Picture (describe the room). And when you are
ready to come back into the room, open your eyes. Welcome back.

Teacher: Okay, any comments?

Student 1: It was quite difficult to keep it going. I was okay for a
while, and then suddenly I realized that I was off thinking about
something completely different.

Teacher: As if you had left the stream altogether.

Student 1: Yeah. I'd forgotten about the exercise.

Teacher: Did you notice what had been happening right before you forgot
about the exercise.

Student 1: Well, I was wondering how this exercise was going to help me
in the summer exam, and this was written on a leaf. But then I started
thinking about what would happen if I failed an exam, and I couldn't go
to the States to work during the summer. I was thinking about how I'd
survive financially next year if that happened, and the next thing I
think is, 'Oh no, I forgot to put all that on a leaf.'

Teacher: That's great. When the thought first showed up, "What if I fail
an exam next summer," was that thought written on a leaf?

Student 1: I'm not sure, I don't think it was.

Teacher: Where was it instead?

Student 1: No where in particular. I was just thinking it.

Teacher: In other words, the thought was located HERE and NOW, about
events next summer, THERE and THEN. Can we say it this way? At some
point you engaged in a piece of private verbal behavior (i.e., you had a
thought) and you failed to respond to it as located THERE and THEN. The
thought was located HERE and NOW, and thus it became "ontologically
true," in the sense that you believed that you really could fail an exam
next summer. So you start actually working out what might happen, what
you will do, and so on, and at that point the stream and the leaves are
gone. The thought is no longer an "ontologically truthless" event that
you respond to dispassionately as it floats down the stream. Instead you
are dealing with the "ontological truth" of exam failure and financial
problems next year. The thought is happening HERE and NOW about events
THERE and THEN.

Student 1: Yeah, I see what you mean.

Teacher: Did you finally get that "exam-failing" thought on a leaf?

Student 1: When I realized that I'd forgotten about the exercise, I put
myself back on the bank, and put the thought on a leaf and watched it
float by.

Teacher: Did things go okay after that?

Student 1: Yeah--sort of. Every now and then, I'd realize that I'd
forgotten about the exercise, and was off thinking all sorts of stuff,
so I'd put myself back on the bank and put the thought that distracted
me on a leaf and watch it float clown the stream.

Teacher: So you were sort of cycling between responding to your thoughts
as THERE and THEN, on the leaves and therefore neither true nor false
ontologically, and HERE and NOW when you got carried off by the thought
and it therefore became ontologically true?

Student 1: Yeah--that's about it.

Teacher: Good. Thanks for sharing that with us. Any more comments from
anyone else?

Student 2: I started out okay, but then I started to think, "This is
really strange--this isn't what I expected in behavioral psychology. I
wonder, if this is some sort of behaviorists' joke."

Teacher: And did you respond to that thought as THERE and THEN, on a
leaf, or did you respond to it as HERE and NOW by getting carried off by
the thought?

Student 2: I got carried off by the thought.

Teacher: What happened to the stream?

Student 2: It disappeared.

Teacher: Right. Isn't it the case that every time the stream
disappeared, it was because you responded to a thought as located HERE
and NOW, rather than located THERE and THEN. In effect, as soon as you
responded to a thought as ontologically true, rather than as just
another thought, you immediately got carried off by the "ontological
truth" of that thought.

Student 2: Yes. That's what seemed to happen.

[Process some of the other students' comments in a similar way, and then
round up with the following]

Teacher: Okay. Thanks for sharing your experiences with the group. That
was great. The point of the exercise was to get you to experience the
difference between responding to a thought as neither true nor false
ontologically, located THERE and THEN, and responding to a thought as
ontologically true, located HERE and NOW (i.e., when you forgot about
the exercise and were carried off by the "ontological truth" of the
thought).

In the same way that the verbal tail chasing undermined ontological
truth for the behavior analyst, the floating leaf exercise can help you
perceive your thoughts about yourself and the world as not ontologically
true. In fact, this can be very useful. The next time you feel anxious
or down about something use the exercise. For example, if you find
yourself thinking, "I'm not clever enough to pass this course, I might
as well give up now," try putting that thought on a leaf, and see how
your reaction to the thought changes. If it works for you, the
ontological truth of the thought will diminish, and you will feel less
inclined to "give up."

Notice also, that if the exercise works for you, you may be inclined to
think of the exercise as true in some ontological sense. If this
happens, then put that thought on a leaf and see how the ontological
truth of the exercise itself diminishes. This will help you see that the
exercise is true only insofar as it works for you --remember the
pragmatic truth criterion?

By the way, I haven't met anyone who can let the leaves float by 100
percent of the time. That is not realistic. With practice, however, you
can let them float by for longer periods of time. You can practice this
at home and you'll be able to do it more and more--it is a kind of
meditative exercise really. In fact, if you practice the exercise
regularly, you may find that it sometimes brings on a state of deep
relaxation, or a feeling of peace and tranquillity. One might even say
that it can lead to a mystical or spiritual experience. This is not to
say, however, that the exercise is "magical." Even the mystical can
yield to a behavior analysis. Consider the following.

The mystical is often spoken of as a nonphysical experience that in some
way transcends physical reality. The mystical experience cannot be
pointed to or defined as a thing with specific boundaries in time and
space, and cannot easily be described in ordinary language. Relational
frame theory accounts, in part, for the transcendental and
nondescriptive qualities of the mystical in terms of a massive reduction
in the evaluative functions actualized by relational framing activities.
Normally, when we engage in relational framing we constantly evaluate,
HERE and NOW, events located THERE and THEN--for example, these objects
are big, small, and medium sized; she is nice and he is nasty; that is
difficult, but that is easy; that will take a long time, but that will
not; and so on. If, however, the stimulus functions actualized by this
activity are dramatically reduced, our evaluations may simply be
observed HERE and NOW as just thoughts or behavioral events located
THERE and THEN. In effect, our "normal" perception of reality, as
ontologically real, is challenged when we experience our own
ever-changing stream of evaluations as evaluations rather than as
reflections of the way things really are. And of course, this experience
of our own evaluations cannot be expressed in ordinary language without
further evaluations, as "good," "bad," "weird," etc. From a relational
frame perspective, this inability to verbalize such an experience,
without further evaluations, gives rise to the intangible and
inexpressible quality of the mystical experience. One is left, so to
speak, with nothing more to say, because anything said would simply be
another evaluation.

Something close to this experience may happen during the exercise as you
place each thought on a leaf and watch it float away down the stream. In
effect, you are learning to respond to your own "inner dialogue" as
neither true nor false, good nor bad, pleasant nor unpleasant,
etc.,--but to simply observe your thoughts as thoughts, and no more.
This, I would argue, is an important feature of the mystical experience,
and it is one that may be reached within the world view we call
behavior-analysis.

[Some further discussion of spirituality and the mystical often arises
at this point, and the work of Dymond and Barnes (1997), Hayes, L. J.
(1997), Hayes, S.C. (1984,1995), and Williams (1986) are normally used
as a means of presenting a behavior-analytic perspective on these issues].

Conclusion <#toc>

Students exposed to the foregoing exercise, and a number of others that
we use in our teaching, have all responded very positively in that we
have noticed a marked increase in student interest in behavior analysis,
and a marked reduction in negative comments about the discipline. As
yet, however, we have not analyzed the effects of the exercise
empirically. We could, for example, ask students to complete
questionnaires before and after the exercise to determine whether it
significantly changes their verbal statements about behavior analysis as
an approach to psychology. An empirical analysis, however, is not
necessary in the current context. We did not present the exercise to
"prove" the effectiveness of a particular teaching method, but instead
aimed to demonstrate how a behavior analyst might respond to the
"troublesome" issue of behavioral reflexivity. When confronted with the
reflexive nature of behavior analysis we responded as behavior analysts
in that we (a) analyzed behavioral reflexivity as a behavioral
phenomenon (i.e., the three scenarios), and (b) we applied the analysis
to a practical problem (i.e., attracting students to behavior analysis
with the teaching exercise). In effect, the statement "behavior analysis
is reflexive" led to behavioral events (the analysis and the teaching
exercise) that were subsequently discriminated as "effective action" in
the behavioral streams discriminated as "Dermot Barnes" and "Bryan
Roche." in the current context, therefore, we find behavioral
reflexivity to be a true and welcome feature of behavior analytic
epistemology.

The authors thank Adam Grundt, Linda Hayes, and Steve Hayes, for their
helpful comments on an earlier version of this paper. The first author
also thanks his students, past and present, who contributed in many
important and diverse ways to the ideas expressed in the current work.
In particular, I thank Veronica Cullinan, Simon Dymond, Ian Grey, Olive
Healy, Neil Hegarty, Geraldine Leader, Dave McAuliffe, Rosemary McCabe,
and Mary Vaughan. I dedicate this paper to Sue, my wife and best friend,
in celebration of our tenth wedding anniversary. Correspondence and
requests for reprints should be sent to Detract Barnes, Department of
Applied Psychology, University College Cork, Cork, Ireland.

   1. Oliver and Landfield (1963) argued that any psychological theory
      must be self-referring or reflexive, in the sense of explaining
      the psychologist's own behavior. In the current paper we are
      concerned only with the reflexive nature of behavior analysis, and
      thus we shall use the term behavioral reflexivity.
   2. Separate functions are defined for two main reasons. First, the
      history of behavioral interactions that established the
      discriminative function of the red light for the response of one
      of the drivers, cannot be identical to the history that
      established the function for the response of the other driver
      (e.g., one driver may have stopped previously at the traffic
      light, whereas the other may not). Insofar as a discriminative
      function is explained, in part, by the history of behavioral
      interactions that produced that function, then clearly the
      functions for the two drivers cannot be considered identical.
      Second, the discriminative function of the red light may be
      manipulated independently in one or other of the behavioral
      streams. Imagine, for example, that a passenger traveling with one
      of the drivers says; "Ignore the lights, I heard a report on the
      radio saying that they are all stuck on red." In this case, the
      discriminative function of the red light may suddenly change in
      one behavioral stream, but not in the other (e.g., one driver will
      drive through the red light, while the other looks on in horror).
   3. To define truth as "useful in achieving certain goals" is to
      define truth behaviorally (e.g., having a particular goal and
      trying to achieve it are behavioral events). The truth of a
      behavior-analytic statement must, therefore, be defined within a
      particular behavioral stream, and as such a truth statement is
      always inherently historical and context specific. From this
      perspective, ontology is simply irrelevant because the behavior
      analyst has no grounds on which to speak of hidden essences and
      underlying realities. That does not mean that ontological talk
      must be abandoned, however, because it may be useful at times to
      speak ontologically. With that caveat, the pragmatic behavior
      analyst takes the view that we cannot take ontological talk (or
      any talk) literally as it applies to an underlying philosophy of
      science or an underlying reality, because truth is just successful
      working--no more and no less.
   4. The idea that a behavior analyst is unable to say anything that
      corresponds to an ontological reality may be seen by some as an
      unbearable conceptual or even psychological burden. In other
      words, if even scientific statements are behavioral events, and as
      such are not ontologically true, what is the point in doing
      science? Interestingly, however, Skinner (1974) did not appear to
      be overly concerned with this issue when he wrote: "If human
      behavior is as fully determined as the behaviorist says it is, why
      does he bother to write a book? Does he believe that anything
      matters? ... Similar questions might as well be asked of the
      author of a book on respiration: 'If that is respiration, why do
      you go on breathing?'" (pp. 247-248).
   5. The conceptual and empirical details of relational frame theory
      have been outlined elsewhere, many times before (see, for example,
      Barnes, 1994; Barnes, 1996; Barnes & Holmes, 1991; Barnes & Roche,
      1996; Hayes, S.C., 1991, 1994; Hayes, S.C., & Hayes, L. J., 1989,
      1992; Hayes, S.C., & Wilson, 1993, 1994, 1996; Lipkens, Hayes,
      S.C., & Hayes, L. J., 1993). Furthermore, detailed knowledge of
      the theory is not required to understand the current
      interpretation, and thus we will not trudge over well-worn ground
      in the current paper.
   6. The following analysis in terms of the relational frames of HERE
      and THERE, and NOW and THEN obviously focuses on four words used
      in the English speaking verbal community. Of course, responding in
      accordance with these frames is also found in many other language
      communities who use different words with similar functions (e.g.,
      ICI in French is functionally equivalent to HERE). Nevertheless,
      such responding may not be universal. The Hopi, for example,
      appear to respond in accordance with frames that may be described
      as HERE and STATIC, and NOW and EMERGING (see Whorf, 1956).
   7. When the behavior analyst utters a pragmatic truth statement, HERE
      and NOW, in relation to an event, THERE and THEN, it may appear as
      if he or she is discriminating a correspondence between the
      statement and the event. However, if the behavior analyst is asked
      whether the statement does, in fact, reflect or correspond to the
      event he or she may simply reply; "I don't know, and I don't care
      - I found making the statement useful, and that is all that matters."
   8. Rejecting behavioral reflexivity does not necessarily require that
      one accept ontological truth. In effect, an argument could be made
      for taking both options (reject behavioral reflexivity and
      ontological truth). A detailed examination of this argument,
      however, is beyond the scope of the current paper (but see Barnes
      & Roche, 1994). Furthermore, in the next section we argue for the
      acceptance of behavioral reflexivity which clearly rules out
      taking both options. Parenthetically, we assume that no one would
      seriously argue for accepting both behavioral reflexivity and
      ontological truth, insofar as they clearly represent contradictory
      positions.
   9. The exercise presented in the current paper is a modified version
      of a technique used in Acceptance and Commitment Therapy for
      undermining excessive rule-following (Hayes, S.C., Strosahl, &
      Wilson, in press). The Teacher-Student dialogues presented in this
      section are based on recordings made during the course "Conceptual
      Issues in Behavior Analysis" run by the first author at University
      College Cork.

552n1.jpg/Figure 1. A diagrammatic representation of a child responding
in accordance with the relational frames of HERE and THERE, and NOW and
THEN. The child's perspective is always located HERE and NOW, and
"reality" is always located THERE and THEN./

554n1.jpg/Figure 2. Upper panel: A diagrammatic representation of the
behavior analyst locating ontological truth statements HERE and NOW that
correspond to a subject's behavior, located THERE and THEN. Middle
panel: Ontological truth statements, as examples of scientific behavior,
shift to THERE and THEN in the relational frames. Lower panel: The
statement, "Scientific statements cannot correspond to reality" is
located HERE and NOW, and corresponds to a lack of correspondence
between what is said and reality, located THERE and THEN. This is the
only remaining ontological truth./

556n1.jpg/Figure 3. Upper panel: The relational functions of THERE and
THEN are attached to the term "effective action," and thus the behavior
analyst concludes that any scientific statement, located HERE and NOW,
is ontologically true to the extent that it leads to effective action
(located, THERE and THEN). Lower panel: Effective action is
discriminated as a behavioral event and thus the behavior analyst is
forced to state that "truth is a property of my own behavior."
Relational responding to ontological truth has been restricted to this
single statement, located HERE and NOW, corresponding to truth and
behavior, located THERE and THEN./

558n1.jpg/Figure 4. Upper panel: The behavior analyst states HERE and
NOW "my last statement was behavior," and thus the previous statement
(i.e., "truth is a property of my own behavior") is located THERE and
THEN in the relational frames. Lower panel: The behavior analyst states
HERE and NOW, "And my last statement was even more behavior," and thus
the previous statement "My last statement was behavior" shifts to THERE
and THEN in the relational frames./

559n1.jpg/Figure 5. Upper panel: The behavior analyst states, HERE and
NOW, that behavioral reflexivity, THERE and THEN, is, for example,
"unproductive," or "non-functional." The relational functions of HERE
and NOW are permanently attached to the ontologically true statement
that "behavioral reflexivity is unproductive." Lower panel: The behavior
analyst's repeated failure to locate the truth HERE and NOW in the
relational frames causes this form of relational responding to habituate
and produces a "meditative" event. In effect, the behavior analyst stops
locating the ontological truth HERE and NOW in the relational frames./

562n1.jpg/Figure 6. Upper panel: The meditative event produces a general
loss of ontological truth and thus the functions of "effective action"
take control. Truth statements are now pragmatic, rather than
ontological. Lower panel: When a pragmatic truth statement is
discriminated as a behavioral event (shifts to THERE and THEN), it
remains true by virtue of its usefulness. In effect, the behavior
analyst discriminates the use of the term "equivalence responding" as
effective action./

563n1.jpg/Figure 7. Upper panel: Responding to truth pragmatically is
discriminated as a behavioral event (shifts to THERE and THEN). Lower
panel: The behavior analyst discriminates the previous discrimination as
ineffective in the current context./

References <#toc>

BARNES, D. (1989). Behavior-behavior analysis, human schedule
performance, and radical behaviorism. The */Psychological Record/*, 39,
339-350.

BARNES, D. (1994). Stimulus equivalence and relational frame theory. The
*/Psychological Record/*, 44, 91-124.

BARNES, D. (1996). Naming as a technical term: Sacrificing behavior
analysis at the altar of popularity (invited commentary). Journal of the
Experimental Analysis of Behavior, 65, 264-267.

BARNES, D., & HOLMES, Y. (1991). Radical behaviorism, stimulus
equivalence, and human cognition. The */Psychological Record/*, 41, 19-31.

BARNES, D., & ROCHE, B. (1994). Mechanistic ontology and contextualistic
epistemology: A contradiction within behavior analysis. The Behavior
Analyst, 17, 165-168.

BARNES, D., & ROCHE, B. (1996). Relational frame theory and stimulus
equivalence are fundamentally different: A reply to Saunders'
commentary. The */Psychological Record/*, 46, 489-507.

DYMOND, S., & BARNES, D. (1995). A transformation of self-discrimination
response functions in accordance with the arbitrarily applicable
relations of sameness, more-than, and less-than. Journal of the
Experimental Analysis of Behavior, 64, 163-184.

DYMOND, S., & BARNES, D. (1997). Behavior-analytic approaches to
self-awareness. The */Psychological Record/*, 47, 181-200.

HAYES, L. J. (1993). Reality and truth. In S.C. Hayes, L. J. Hayes, H.
W. Reese, & T. R. Sarbin (Eds.), Varieties of scientific contextualism
(pp. 35-44). Reno, NV: Context Press.

HAYES, L. J. (1997). Understanding mysticism. The */Psychological
Record/*, 47, 573-596.

HAYES, S. C. (1984). Making sense of spirituality. Behaviorism, 12, 99-110.

HAYES, S.C. (1991). A relational control theory of stimulus equivalence.
In L J. Hayes & P. N. Chase (Eds.), Dialogues on verbal behavior: The
first international institute on verbal relations (pp. 19-40). Reno, NV:
Context Press.

HAYES, S.C. (1993). Analytic goals and the varieties of scientific
contextualism. In S.C. Hayes, L. J. Hayes, H. W. Reese, & T. R. Sarbin
(Eds.), Varieties of scientific contextualism (pp. 11-27). Reno, NV:
Context Press.

HAYES, S.C. (1994). Relational frame theory: A functional approach to
verbal events. In S. C. Hayes, L. J. Hayes, M. Sato, & K. Ono (Eds.),
Behavior analysis of language and cognition (pp. 9-30). Reno, NV:
Context Press.

HAYES, S.C. (1995). Knowing selves. The Behavior Therapist, 18, (No.3),
94-96.

HAYES, S.C., & BROWNSTEIN, A. J. (1986). Mentalism, behavior-behavior
relations, and a behavior-analytic view of the purposes of science. The
Behavior Analyst, 9, 175-190.

HAYES, S.C., & HAYES, L. J. (1989). The verbal action of the listener as
a basis for rule-governance. In S.C. Hayes (Ed.), Rule-governed
behavior: Cognition, contingencies, and instructional control (pp.
153-190). New York: Plenum.

HAYES, S.C., & HAYES, L. J. (1992). Verbal relations and the evolution
of behavior analysis. The American Psychologist, 47, 1383-1395.

HAYES, S.C., STROSAHL, K., & WILSON, K. G. (in press). Acceptance and
commitment therapy. New York: Guilford.

HAYES, S.C., & WILSON, K. G. (1993). Some applied implications of a
contemporary behavior-analytic view of verbal events. The Behavior
Analyst, 16, 283-301.

HAYES, S.C., & WILSON, K. G. (1994). Acceptance and commitment therapy:
Altering the verbal support for experiential avoidance. The Behavior
Analyst, 17, 289-303.

HAYES, S.C., & WILSON, K. G. (1996). Criticisms of relational frame
theory: Implications for derived stimulus relations. The */Psychological
Record/*, 46, 221-236.

KEENAN, M. (1996). "W"-ing: Private events in the classroom. Behavior
and Social Issues, 6, 75-84.

LEIGLAND, S. (1993). Scientific goals and the context of justification.
In S.C. Hayes, L. J. Hayes, H. W. Reese, & T. R. Sarbin (Eds.),
Varieties of scientific contextualism (pp. 28-33). Reno, NV: Context Press.

LIPKENS, R., HAYES, S.C., & HAYES, L. J. (1993). Longitudinal study of
the development of derived relations in an infant. Journal of
Experimental Child Psychology, 56, 201-239.

OLIVER, W. D., & LANDFIELD, A. W. (1963). Reflexivity: An unfaced issue
of psychology. Journal of Individual Psychology, 20, 187-201.

SCHNAITTER, R. (1987). Behaviorism is not cognitive and cognitivism is
not behavioral. Behaviorism, 15, 1-11.

SCHOENFELD, W. N., & FARMER, J. (1970). Reinforcement schedules and the
"behavior stream." In W. N. Schoenfeld (Ed.), The theory of
reinforcement schedules. New York: Appleton.

SKINNER, B. F. (1938). The behavior of organisms: An experimental
analysis. New York: Appleton-Century-Crofts.

SKINNER, B. F. (1957). Verbal behavior. New York: Appleton-Century-Crofts.

SKINNER, B. F. (1969). Contingencies of reinforcement: A theoretical
analysis. New York: Appleton-Century-Crofts.

SKINNER, B. F. (1974). About behaviorism. London: Jonathan Cape.

SPINELLI, E. (1989). The interpreted world. London: Sage.

STADDON, J. E. R. (1993). Pepper with a pinch of psalt. The Behavior
Analyst, 16, 245-250.

STEELE, D., & HAYES, S.C. (1991). Stimulus equivalence and arbitrarily
applicable relational responding. Journal of the Experimental Analysis
of Behavior, 56, 519-555.

WHORF, B. L. (1956). Science and linguistics. In J. B. Carroll (Ed.),
Language, thought, and reality: Selected writings of Benjamin Lee Whorf.
Cambridge, MA: MIT Press.

WILLIAMS, J. L. (1986). The behavioral and the mystical: Reflections on
behaviorism and eastern thought. The Behavior Analyst, 9, 167-173.

ZURIFF, G. E. (1995). A review of Baum's Understanding behaviorism:
Science behavior, and culture. The Behavior Analyst, 18, 395-400.

~~~~~~~~

By DERMOT BARNES and BRYAN ROCHE, University College Cork