Content uploaded by Sean Hughes
Author content
All content in this area was uploaded by Sean Hughes on Nov 27, 2017
Content may be subject to copyright.
RUNNING HEAD: INTRODUCTION TO SPECIAL ISSUE
The Functional-Cognitive Framework for Psychological Research:
Controversies and Resolutions
Sean Hughes & Jan De Houwer
Ghent University
Marco Perugini
Universita di Milano-Bicocca
Author Note
SH and JDH, Department of Experimental Clinical and Health Psychology, Ghent
University, Belgium. MP, Faculty of Psychology, University of Milan-Bicocca, Milan, Italy.
The preparation of this paper was made possible by Grant BOF09/01M00209 of Ghent
University to JDH and VBO grant 01T02512 of Ghent University to JDH and MP.
Correspondence concerning this article should be sent to sean.hughes@ugent.be.
Abstract
The scientific goals, values and assumptions of functional and cognitive researchers have
propelled them down two very different scientific pathways. Many have, and continue to
argue, that these differences undermine any potential communication and collaboration
between the two traditions. We explore a different view on this debate. Specifically, we focus
on the Functional-Cognitive (FC) framework, and in particular, the idea that cognitive and
functional researchers can and should interact to the benefit of both. Our paper begins with a
short introduction to the FC framework. We sweep aside misconceptions about the
framework, present the original version as it was outlined by De Houwer (2011) and then
offer our most recent thoughts on how it should be implemented. Thereafter, we reflect on its
strengths and weaknesses, clarify the functional (effect-centric vs. analytic-abstractive) level
and consider its many implications for cognitive research and theorizing. In the final section,
we briefly review the articles contained in this Special Issue. These contributions provide
clear examples of the conceptual, empirical and methodological developments that can
emerge when cognitive, clinical, personality and neuroscientists fully engage with the
functional-cognitive perspective.
Keywords: Functional-Cognitive Framework, Meta-theory, Cognitive Psychology, Functional
Psychology
INTRODUCTION TO SPECIAL ISSUE 1
The Functional-Cognitive Framework for Psychological Research:
Controversies and Resolutions
Stop for a moment and imagine that you are a relative newcomer to the world of
psychological science. Stretched before your eyes would lie an archipelago of academic
islands, each home to a unique “tribe” or sub-discipline like social and cognitive psychology,
as well as clinical, personality and neuropsychology. Travelling from island to island you
would find that each tribe has devised its own unique culture (theories), language
(terminology), set of tools (procedures) and (analytic) practices. Although these tribes
typically operate independently from one another, most of them are united by a shared
adherence to a believe system (philosophy of science) known as the mental mechanistic
system. At its core resides three simple ideas. The first is that changes in behavior occur when
organisms interact with the environment. Second, that these changes are mediated by
mechanisms, that is, processes that are themselves composed of discrete parts that interact
with one another and are subject to specific operating conditions (Bechtel, 2008). Third, these
mechanisms are mentalistic in nature. Therefore the scientists’ goal is to develop and test
theories about the mental processes and representations via which organisms’ store, process
and retrieve the information assumed to influence their behavior (see Bechtel, 2008; De
Houwer, 2011).
For well over forty years now the mental mechanistic approach has influenced the
scientific values, goals and assumptions of many, if not most in psychological science.
Although it has undoubtedly accelerated our understanding of the human mind (for reviews
this work see Eysenck & Keane, 2000; Miller, 2003; but see Fiedler, 2014, this issue), it is
important to realize that it is not the only way of studying psychological phenomena. Indeed,
elsewhere in the archipelago one would find a number of tribes that are also studying the
origins and properties of human behavior, language, and cognition. Setting foot on one of
INTRODUCTION TO SPECIAL ISSUE 2
these islands, you would encounter a group of functional researchers who have developed
their own belief system (behaviorism) and crafted a culture (theory), language (terminology),
set of tools (procedures) and practices that differ to those seen in psychological science.
Rather than hunting for mediating mental mechanisms, these functional researchers first set
out to discover which elements in the (past and present) environment moderate changes in
behavior. They then use this information to develop abstract types of functional knowledge
(“laws of behavior”) and eventually theories which enable them to predict-and-influence the
phenomena of interest (see Chiesa, 1994; Zettle, Hayes, Biglan, & Barnes-Holmes, in press,
for book length treatments)
1
.
The Functional-Cognitive Framework
It should come as no surprise that these two beliefs systems (mental mechanistic vs.
behaviorism) have propelled their proponents down different scientific pathways, each with
its own ideas about the value of certain types of theorizing and empirical findings (e.g., Hayes
& Brownstein, 1986; Gardner, 1985). Unable to reconcile these differences, the two traditions
have largely gone their separate ways, sometimes fighting over their perceived scientific
legitimacy, and more often than not, ignoring the fruits of their respective labors (see Brown
& Gillard, 2015; Baron-Cohen, 2014; Miller, 2003). There have been many excellent
treatments of this topic over the years and Roediger (2004) offers a particularly good one. In it
he refers to an exchange with a colleague (Endel Tulving) who argues that “psychology now
designates at least two rather different sciences, one of behavior and the other of the mind.
They both deal with living creatures, like a number of other behavioral sciences, but their
1
Note that for the sake of communication we have somewhat simplified our story. For instance, few people
realize that there are many different varieties of behavioral thinking that differ at both the level of philosophy
and scientific practice (e.g., methodological behaviorism, radical behaviorism, contextual behavioral science).
Instead, and more often than not, the term behaviorism is (incorrectly) equated with a mechanistic rather than a
functional type of thinking, albeit one that is based on behavioral (e.g., S-R) rather than mental mechanisms.
Rather than unpack this conceptual issue here, we will simply refer to “functional researchers” from now on
whenever we want to speak of that broad group who conduct their scientific work at the functional level of
analysis, that is, at the level of relations between environment and behavior.
INTRODUCTION TO SPECIAL ISSUE 3
overlap is slim, probably no greater than psychology or sociology used to be when the world
was young. No one will ever put the two psychologies together again, because their subject
matter is different, interests are different, and their understanding of the kind of science they
deal with is different. Most telling is the fact that the two species have moved to occupy
different territories, they do not talk to each other (any more), and the members do not
interbreed. This is exactly as it should be.”
The special issue that you are now reading explores a different view on this debate. It
focuses on the Functional-Cognitive (FC) framework, and in particular, the idea that cognitive
and functional researchers can and should interact to the benefit of both (De Houwer, 2011).
Simply put, the FC framework indicates that psychological science can be carried out at one
of two mutually supportive levels of analysis: either (a) at the cognitive level where
researchers investigate the mental mediators of a behavioral effect or at (b) the functional
level where researchers investigate the environmental moderators of behavior. When research
is conducted within the remit of this framework, a number of theoretical, methodological and
empirical benefits emerge for both types of researchers. As we will discuss in more detail
below, for those interested in the cognitive level it slams the door shut on those pernicious
problems that follow from treating behaviors-as-proxies, kick starts the discovery of new
mental mechanisms, and can help refine existing mental theories. For their counterparts
interested in the functional level, it reveals a whole new world of procedures, cognitive
theories and findings that can be used to organize existing behavior, predict novel behavior
and orientate researchers towards previously uncharted research domains.
The current paper, and those elsewhere in this special issue, have an admittedly
ambitious goal - to explore the transformative potential of the FC framework for research
conducted within psychological science. We begin with a short introduction to the framework
itself. What has become clear from editing this special issue, speaking at international
INTRODUCTION TO SPECIAL ISSUE 4
meetings and from discussions with colleagues, is that people differ in their perceptions of the
framework’s core arguments and intentions. These perceptions are often colored by historical
assumptions about, and lack of interactions between, cognitive and functional researchers. At
the same time, our own understanding of the framework has also evolved through full and
active engagement with both traditions. We have come to appreciate that although it can be
instantiated in several ways, only one of these instantiations is likely to meet with success.
Therefore, we will first sweep aside any misunderstandings about the framework, present the
original version that was outlined by De Houwer (2011) and then offer our most recent
thoughts on how it should be implemented. Thereafter, we consider several issues that may
influence the decision to adopt the framework and then speculate about its future. The final
section briefly reviews the articles contained in this Special Issue. These contributions
evaluate the framework’s strengths and weakness and highlight the conceptual, empirical and
methodological developments that can emerge when cognitive, clinical, personality and
neuroscientists fully engage with this perspective
2
.
What Exactly is the Functional-Cognitive Framework?
2
Another short note on terminology seems warranted here – especially given that many terms are used
interchangeably across different levels of analysis. When we use the term functional we are referring to relations
between (past and present) environment and behavior and not to the fact that (a) something is adaptive for the
organism or to (b) the function or “purpose” of a mental construct. Likewise, when we use the term cognitive we
are referring to the mental level, and in particular, to “the complete set, rather than a subset, of all mental
processes and representations, including affective and motivational constructs” (De Houwer et al., 2013, p.254).
INTRODUCTION TO SPECIAL ISSUE 5
Figure 1. A visual illustration of the various ways in which the functional-cognitive (FC) framework can be
implemented. A first, misconceived framework, would involve forcing cognitive science and behavior analysis
together. The original version of the framework (De Houwer, 2011) involves highlighting two separate but
mutually supportive levels of analysis. The updated version of the framework makes explicit the distinction
between the effect-centric and analytic-abstractive functional approaches.
Misconceived FC framework. One of the core ideas in the FC framework is that
research can be carried out at two mutually supportive levels of explanation. Some seem to
have interpreted this as an attempt to merge cognitive science and (certain branches of)
behavior analysis. It is therefore important to make explicit that the FC framework was never
intended as a call for the merging of specific intellectual traditions. Such a view misrepresents
the FC framework in two ways. First, the framework does not differentiate between the
cognitive and functional levels on the basis of group membership (i.e., which “tribe” or
scientific tradition one identifies with) but on the basis of scientific goals (i.e., the type of
explanation that one focuses on). Most researchers who call themselves cognitive probably
focus on mental mechanisms whereas those belonging to behavior analytic societies probably
focus on environmental influences on behavior. However, it is important to realize that within
each tradition, people can differ in their scientific goals and that across traditions, there can be
communalities in scientific goals.
Cognitive Level
Functional Level
Cognitive Science
+
Behavior Analysis
Cognitive Level
Functional Level
(Effect-Centric)
Functional Level
(Analytic-Abstractive)
Misconceived F-C Framework
Original F-C Framework
Updated F-C Framework
INTRODUCTION TO SPECIAL ISSUE 6
Second, the idea of merging cognitive science and behavior analysis implies that
researchers would need to give up their scientific goals. The FC framework, on the contrary,
acknowledges that researchers can have different goals and thus different research agendas.
The logic here is as follows: most cognitive scientists are interested in identifying the mental
states and operations that causally mediate between environmental input and behavioral
output. Analyses carried out by their functional counterparts are scientifically unsatisfactory
given that they say nothing about the mental mechanics of the mind. From a cognitive
perspective, functional analyses simply yield incomplete “descriptions” of environment-
behavior interactions that lack a conceptual framework that can heuristically organize existing
data or generating novel predictions. At the same time, functional researchers focus on
environment-behavior relations in an effort to predict-and-influence the phenomenon of
interest. They typically consider analyses that incorporate mental states and operations as
speculative, incomplete, and as potentially distracting from the scientific goal at hand (see
Stewart, this issue). Therefore instantiating the FC framework as a merger of the two
approaches will actually hamper rather than optimize scientific progress because it fails to
reconcile the different scientific agendas of cognitive and functional researchers. In fact, such
a misconceived framework actually rejects the idea that research carried out in these two
traditions can be mutually supportive because a denial of the fundamental difference in goals
would imply that both approaches are striving for the same things and are thus competing for
scientific legitimacy.
Original FC Framework. Let us be clear here. The FC framework outlined by De
Houwer (2011) bears little resemblance to that described in the previous section. Instead, the
framework distinguishes two different levels of explanation that vary in their explanatory
targets and constructs. Functional explanations are directed at explaining behavior in terms of
environmental events. Cognitive explanations are directed at explaining environment-
INTRODUCTION TO SPECIAL ISSUE 7
behavior relations in terms of mental mechanisms. There are different reasons why some
researchers would prefer functional explanations (e.g., because they have the aim to predict-
and-influence behavior) whereas others prefer cognitive explanations (e.g., because for them
something is “truly” understood only if the mechanism is understood). The FC framework
does not interfere with the goals of a researcher, not does it pass judgment on those goals or
the reasons behind those goals. Instead, it emphasizes that both levels of explanation can be
mutually supportive for one another. By conceptualizing the two levels in this way, the
framework purposefully sets debates about the scientific primacy of intellectual traditions or
scientific approaches to the side (see Reyna, 1995, for such discussions). It does not see
differences in explanatory targets and concepts as an insurmountable obstacle to
communication between the two levels nor does it view these differences as inevitably leading
to competition between intellectual traditions. Rather it draws upon the knowledge gained at
one level to advance progress at the other.
Merits of the original FC framework. We believe that the original FC framework had
many merits. First, it highlighted the functional level of analysis, and by implication, a way of
explaining human behavior that is important in its own right. Unlike the cognitive level,
functional analyses do not appeal to, or make any assumptions about, mental states and their
operation. Instead they involve the identification of functional relations between behavior and
the environment (i.e., behavioral effects). These relations are not descriptive but rather
explanatory: establishing the presence of an effect “implies a hypothetical explanation of the
behavior that is couched in terms of elements in the environment rather than mental
constructs” (De Houwer, 2011, p.205).
Second, by drawing attention to the functional level and requiring outcomes to be
defined as behavioral effects rather than mental constructs, the framework offered researchers
a practical way to side-step the various problems that plague the behavior-as-proxy approach.
INTRODUCTION TO SPECIAL ISSUE 8
Simply put, the behavior-as-proxy problem refers to the fact there appears to be no method
with which to directly interact with mental constructs. Researchers are constrained insofar as
they can only act on the world in some way, observe a change in behavior, and based on the
effect of their actions, postulate that a mental mechanism (operating under a set of mental
conditions) is responsible for the obtained outcome. Those subscribing to the mental approach
often attempt to circumvent this problem by treating a behavioral effect as a “proxy” for the
mental construct under investigation. In other words, the presence of a particular change in
behavior - and the environmental conditions under which it is observed - is treated as
evidence for the presence of a mental construct and the conditions under which it operates.
The problem is that this approach is built on questionable assumptions (e.g., that the mental
process is the only determinant of the behavioral effect). Violations of these assumptions can
undermine the construction of mental theories and the interpretation of empirical data (see De
Houwer, 2011; De Houwer, Gawronski, & Barnes-Holmes, 2013). Unfortunately, treating
behavior as a proxy for mental constructs is not an isolated practice but rather an all too
common one in psychological science (see Eagly & Chaiken, 2007; Fazio, 2007; Krosnick,
Judd, & Wittenbrink, 2005; Poldrack, 2006 for similar arguments). The FC framework draws
attention to the above problems, equips psychologists with a means to avoid conflating
behavioral effects and mental constructs, and does so in a way that is applicable to a wide
range of psychological phenomena.
A third merit of the FC framework is that it emphasized that while the cognitive and
functional levels are independent from one another they are still deeply intertwined.
Researchers operating at the cognitive level often formulate their mental theories by drawing
upon existing environment-behavior relations (behavioral effects) and test the utility of those
theories by generating new effects. By accumulating and refining their understanding of these
effects they constrain existing mental models and force them to increase the precision of their
INTRODUCTION TO SPECIAL ISSUE 9
assumptions or even to alter them altogether. Thus activity at the functional level provides the
raw fuel (behavioral effects) needed to drive the engines of mental theorizing. In this way, a
strong functional approach is a basic requirement for a strong cognitive approach (for similar
arguments see Fiedler, this issue). At the same time, the framework also highlighted that
theories at the mental level can systematically organize functional knowledge (i.e., heuristic
value) and generate new hypotheses about the conditions under which those effects occur
(i.e., predictive value). In this way a strong mental approach can also facilitate a strong
functional approach.
Finally, the framework clarified that this beneficial interplay is only going to work
when the two levels are strictly separated from one another at a conceptual level. This means
that (a) separate terms should be used to explain behavioral effects and mental constructs, (b)
terms from different levels should not be intermixed, and that (c) empirical findings should
first be described, as much as possible, at the functional level before doing so at the cognitive
level. This ensures that a priori ideas about mental mechanisms do not constrain the number
and type of constructs that are considered relevant for an effect, or the effects relevant to those
constructs. This approach increases the speed of theoretical innovation (by removing
restrictions on which constructs are considered relevant to the effect), facilitates the discovery
of new functional knowledge (by allowing for alternative mental theories and thus novel
predictions), and ensures that the accumulated body of functional knowledge remains valid
regardless of changes in mental theories across time. Therefore, just as certain practices
increase the likelihood of replicating psychological research (Nosek et al., 2015), the FC
framework increase the likelihood of open-mindedness and cumulative growth when carrying
out that research. To some extent, the merit of the framework is evident from the fact that it is
already beginning to reshape how we think about a variety of domains, from attitudes (De
Houwer et al., 2013a), and learning (De Houwer et al., 2013b) to evaluative conditioning
INTRODUCTION TO SPECIAL ISSUE 10
(Hughes, De Houwer, & Barnes-Holmes, in press), cognitive control (Liefooghe & De
Houwer, this issue), personality (Perugini, Costantini, Hughes, & De Houwer, this issue),
neuroscience (Vahey & Whelan, this issue) and clinical psychology (De Houwer, Barnes-
Holmes, & Barnes-Holmes, this issue). Indeed, thumbing through the pages of this special
issue reveals that the above arguments translate into concrete recommendations for scientific
activity in a wide variety of psychological domains.
Reflecting on the framework. Despite its many merits, the original FC framework was
obviously not perfect. One could argue that it offered nothing that cognitive scientists were
not already doing, insofar as grafting a functional level (focused on behavioral effects) onto a
mental level is simply a way of relabeling (good) cognitive psychology. Surely all
experiments involve a functional approach wherein independent variables are manipulated
and their effects on dependent variables examined? Likewise, the framework did not require
any interaction between cognitive and functional researchers so long as the former clearly
separated their mental constructs from their behavioral effects. By equating the functional
level with behavioral effects the framework may have presented an overly simplified,
unrepresentative view of that level which did not exploit its true potential (Zettle et al., in
press; Hughes & Barnes-Holmes, in press-a). Thus interactions between cognitive scientists
and functional researchers were optional at best, and unnecessary at worst (for related
arguments see Barnes-Holmes & Hussey, this issue).
Clarifying the FC Framework. Taking a step back, we believe that the framework
can be developed further by not only addressing the above issues but by also unleashing the
full potential of the functional level. This can be achieved by highlighting that there are two
ways of conducting research at the functional level: either in an effect-centric or analytic-
abstractive manner. The effect-centric approach is what we have discussed to this point and is
probably what most psychologists are familiar with: it involves manipulating an independent
INTRODUCTION TO SPECIAL ISSUE 11
variable (environment) and measuring its effects on a dependent variable (behavior). For
instance, researchers might compare children who are told that access to ice-cream depends
on their eating vegetables at dinner to child who receive no such instructions and then
measure their respective vegetable consumption. They might also examine whether use of
Facebook increases or decreases when people receive content alerts, or even if the likelihood
that a dog will bite the postman changes following one type of obedience training versus
another. Although these analyses generate precise functional knowledge about the relationship
between environment and behavior, it is a type of knowledge that is limited in its scope (i.e.,
effects that only apply to a certain procedure or sets of situations).
It is important to realize that a second functional approach also exists. This analytic-
abstractive approach is typically used in behavior analysis and involves a two-step process of
(a) identifying specific functional relations between environment and behavior, and then (b)
abstracting these relations into general behavioral principles that are precise (explain a
specific set of behaviors), far reaching (explain a comprehensive range of behaviors across a
variety of situations) and scientifically coherent (consistent across analytical levels and
domains such as biology, psychology, and anthropology; see Barnes-Holmes & Hussey, this
issue; Hughes et al., in press). Examples of these principles include reinforcement and
stimulus control (see Liefooghe & De Houwer, this issue). Functional researchers seek this
type of abstract functional knowledge out because it can reveal similarities and differences
between effects that would otherwise remain hidden. Moreover, by weaving different
behavioral principles together, they can and have developed functional theories. These
theories allows them to predict-and-influence a wide range of psychological phenomena
including language, problem solving, implicit cognition, self and perspective taking,
intelligence and psychopathology (for recent reviews see Hughes & Barnes-Holmes, in press-
a,b; Stewart, this issue). Please note that these two functional approaches (effect-centric and
INTRODUCTION TO SPECIAL ISSUE 12
analytic-abstractive) can be seen as two end-points of a continuum ranging from
topographical (i.e., formulated in terms of superficial characteristics of dependent and
independent variables that apply only to one or limited set of variables) to abstract (i.e.,
formulated in terms of abstract characteristics of dependent and independent variables that
apply to a wide range of variables).
Merits of the analytic-abstractive level. The analytic-abstractive level was certainly
alluded to, but not articulated in, the original FC framework. By actively incorporating it here
we highlight a new functional level with its own unique merits. Foremost amongst these is the
idea of abstract functional knowledge that is not restricted to a certain procedure but that can
explain a wide variety of topographically different outcomes (i.e., knowledge that is both
precise and far reaching). Access to this type of knowledge rarely (if ever) emerges within a
single line of effect-centric research. To illustrate why, take the previous example of
children’s eating habits, a disobedient pet and persistent checking of one’s Facebook profile.
Approaching these behaviors in an effect-centric manner might lead researchers to postulate
three separate effects (e.g., a dessert effect, a disobedience effect, and a checking effect) and
suggest that, because those behaviors look different, involve different organisms, stimuli and
events, they must reflect three separate and unrelated phenomena. Yet from an analytic-
abstractive position these three behaviors can all be viewed as instances of the same
phenomenon (reinforcement). Similarly, stopping one’s car at a red traffic light and
accelerating in the presence of a green light or taking a cake out of an oven after an alarm
rings can both be explained in terms of stimulus control (see Liefooghe & De Houwer, this
issue).
This analytic-abstractive level has many implications for the FC framework. As we
have seen, scientific analyses rarely terminate once a behavioral effect and its moderators
have been identified. Rather researchers take this basic functional knowledge (effects) and
INTRODUCTION TO SPECIAL ISSUE 13
operate on it in some way to better understand, predict and/or influence the phenomena of
interest. The revised FC framework highlights that this “interpretative” operation can occur
not only at the cognitive level as is often assumed (i.e., by postulating mental mechanisms
that mediate between environment and behavior) but also at the functional level (by
abstracting effects into behavioral principles and by weaving these principles together to
create functional theories). It also suggests that the type of interpretive operation applied to
behavioral effects will likely depend on the one’s goals, values and assumptions. Regardless
of this decision, it is important to realize that progress at the functional (analytic-abstractive)
level can lead to progress at the cognitive level and vice-versa.
Consider, for example, the cognitive level. Although an effect-centric approach solves
the behavior-as-proxy dilemma by conceptually separating to-be-explained effects and
explanatory mechanisms, it runs the risk that research agendas become overly narrow and
fragmented. Effect-centric functional research tends to focus on whether and when a specific
effect occurs and less on the possibility that this effect is just one of many different instances
of a given phenomenon. Cognitive research on the mental mediators of the effect also tends to
focus on the mental processes underpinning a specific effect and less on process accounts that
apply across many different domains
3
. This approach differs dramatically from the analytic-
abstractive position which provides a comprehensive, unifying way to describe many different
effects in non-mental terms. Once cognitive researchers adopt an analytic-abstractive
functional perspective, novel questions can emerge about the mental processes that are
common to those different effects. To illustrate, take the topic of cognitive control. In this
domain the Stroop effect, Gratton effect, and Simon effect all refer to the specific outcomes of
particular procedures and are often thought to be underpinned by different mental
3
Note that this is typically but not always the case. For instance, dual-process (Gawronski & Creighton, 2013)
and overarching cognitive theories that apply across a number of domains have attempted to explain a wide
variety of effects (e.g., Anderson, 2013).
INTRODUCTION TO SPECIAL ISSUE 14
mechanisms. As Liefooghe and De Houwer (this issue) demonstrate, these and many other
cognitive control effects may actually (functionally speaking) represent instances of the same
behavioral principle (stimulus control). Such an approach offers much that an effect-centric
position does not. It sets the stage for a general taxonomy of cognitive control wherein effects
are organized according to their functional commonalities and differences. It gives rise to a
conceptual platform that facilitates communication between cognitive researchers who adopt
different terminologies and mental theories. It offers a framework for creating new tasks,
discovering new effects, and by implication, unlocking new questions about cognitive control
processes (see Perugini et al., this issue for similar arguments in the context of personality).
And finally, it highlights the possibility that topographically different but functionally similar
effects might be mediated by similar mental processes.
In other words, the analytic-abstractive level not only unlocks new ways of
conceptualizing data at the functional level but also provides another type of input for
building and evaluating mental theories. Just as basic functional knowledge (effects) provides
the input for specific mental theories so too can abstract functional knowledge (behavioral
principles) provide the input for overarching mental theories. This abstract functional
knowledge places no a priori constrains on mental theorizing given that it has nothing to say
about the mental mediators of behavioral principles. But it does place a posteriori restrictions
on theories by highlighting consistencies across outcomes that need to be explained (e.g., why
is it that behavior controlled by its consequences regardless of the stimulus, procedure or
organisms involved?). This abstract knowledge can also be used to constrain the rate of
theoretical expansion at the mental level and provide another dimension along which existing
theories are evaluated. For instance, if the value of a mental theory is (in part) determined by
its heuristic and predictive value (Gawronski & Bodenhausen, 2014) then overarching mental
models that can explain entire laws of behavior (rather than single effects) are going to have
INTRODUCTION TO SPECIAL ISSUE 15
greater heuristic and predictive power than those that cannot. Thus a strong functional
(analytic-abstractive) level can lead to a strong cognitive level.
Incorporating the analytic-abstractive level also has merits for empirical research. If it
is the case that an effect is just one instance of a behavioral principle then all knowledge
accumulated about that principle can immediately be applied to an effect. Take our previous
example of vegetable consumption as an instance of reinforcement wherein the probability
that the child will eat their greens depends on its consequences (e.g., access or removal of a
dessert). Over fifty years’ worth of functional knowledge indicates that the relationship
between behavior (vegetable eating) and its consequences is influenced by many factors, from
the nature of the context (having dinner at home vs. at a restaurant), and stimuli (e.g.,
presence vs. absence of a dessert menu), to the responses (e.g., crying vs. eating vegetables),
consequences (access to dessert vs. reprimand), and organisms involved (e.g., parent vs.
babysitter) (see Catania, 2007). Similarly, if it is the case that cognitive control effects are just
one instance of a larger principle known as stimulus control then the above argument also
apply (i.e., our understanding of the moderators of stimulus control can be used to inform our
understanding of cognitive control effects; Liefooghe & De Houwer, this issue). This
argument has recently been applied to, and may reshape our thinking in, other domains such
as evaluative conditioning (Hughes, De Houwer, & Barnes-Holmes, in press).
It is worth pausing to reflect that the analytic-abstractive level not only provides
insight into existing effects but can also stimulate entirely new empirical discoveries. Take the
application of the framework to the study of attitudes (De Houwer et al., 2013). In this paper
De Houwer and colleagues argued that attitude research involves the study of evaluation,
which in turn is defined as the impact of stimuli on evaluative responses. These responses are
usually established in one of four ways: either by manipulating regularities in the presence of
a single stimulus (e.g., mere exposure), two or more stimuli (evaluative conditioning),
INTRODUCTION TO SPECIAL ISSUE 16
between behavior and its consequences (approach/avoidance learning), or by instructions
(persuasion). Building on this suggestion, Hughes, De Houwer, and Perugini (in press)
identified a fifth and previously undiscovered way of changing evaluative responses (i.e., via
intersections between regularities). The discovery of this pathway was a direct consequence
of applying the FC framework to the domain of attitudes. Thus a strong functional (analytic-
abstractive) level can lead to a strong empirical developments at the effect-centric (functional)
and mental levels.
The benefits of the revised framework are by no means unidirectional. For those
operating at the analytic-abstractive level the theories and procedures devised, as well as
findings obtained at the mental level could serve an important “orientating” function. In other
words, developments at the mental level could highlight previously undiscovered domains
with significant implications for human behavior as well as new procedures for capturing and
manipulating those behaviors. Indeed, the cognitive literature is replete with phenomena that
have yet to be systematically explored in analytic-abstractive terms, such the behaviors people
refer to when they use words such as creativity and imagination, intelligence, persuasion,
obedience to authority, judgment and decision making, emotional and moral development,
close relationships and personality. While recognizing that not all phenomenon identified at
one level will necessarily be of interest to those operating at the other (Barnes-Holmes &
Hussey, this issue), we do believe that there are many aspects of human psychological life that
have been studied at the cognitive level which would also be of interest to those at the
functional level. Several authors have already adopted such a perspective, noting how
procedures like the IAT and mental theories of automatic evaluation have influenced their
own thinking and activity at the functional level (e.g., Barnes-Holmes, Barnes-Holmes,
Stewart, & Boles, 2011). Therefore a strong cognitive level can lead to a strong functional
(analytic-abstractive) level by orientating researchers towards novel research domains,
INTRODUCTION TO SPECIAL ISSUE 17
equipping them with procedures to explore those domains, and highlighting basic functional
knowledge (effects) that has already been accumulated in those domains.
Challenges Facing the FC Framework
Despite these merits we see reasons why readers may still be reluctant to apply the FC
framework to their own work. First, a strict cost-benefit analysis may lead them to the
conclusion that it requires a high initial “buy-in” with an uncertain potential return. Extracting
the full benefit of the framework requires that they become comfortable with a foreign
(mental or functional) language and at least appreciate the scientific assumptions, values and
goals of their colleagues operating at different levels of analysis. Second, there may be an
inherent lack of motivation for cognitive and functional (analytic-abstractive) researchers to
interact given that their scientific agendas lead them down two very different pathways.
Certain questions and outcomes may serve to stimulate those working at one level and yet be
entirely obvious or redundant to those working at another level (see Barnes-Holmes &
Hussey, this issue). Third, cognitive researchers may certainly recognize the need to
conceptually separate constructs from effects but fail to see how the framework could be
implemented in their research domain or how it can improve their own scientific activity.
Fourth, researchers operating at the analytic-abstractive functional level may see little value in
the framework at all. They already have ways of interpretatively operating on behavior (i.e.,
principles and theories) without the need to recourse to mental mechanisms. They could even
argue that there is an asymmetric utility built into the framework insofar as one (mental) level
needs the other (functional) but the second does not need the first (also see Fiedler, this issue).
Therefore, unless the framework tangibly contributes to scientific activity at these two levels,
then researchers may certainly stop and consider it, but then continue on with business as
usual.
INTRODUCTION TO SPECIAL ISSUE 18
At this moment in time, we can respond to these doubts only by reaffirming our belief
that adopting the framework can lead to unique benefits for both cognitive and functional
researchers, benefits that outweigh the potential costs. But we realize that ultimately, the
value or “success” of the FC framework will be judged on its ability to stimulate empirical,
theoretical and methodological progress in different areas of psychological science in the
years to come. Our recent work, and that outlined in this special issue, provide just a taste of
what can be achieved when the framework is applied to the study of attitudes, learning,
evaluative conditioning, cognitive control, psychotherapeutic strategies, neuroscience and
personality. But this is really only the tip of the iceberg. The framework could be taken by
others and applied to topics such as motivation, attention, memory, judgments, decision
making or indeed any other research domain. The FC framework can function as a bridge
between different islands in the psychological archipelago. Knowledge accumulated at one
island can now flow to the other without reverting to ideological debates about which island
has the best belief system, culture, language or practices.
The Special Issue
The first paper of the special issue (Stewart) focuses on the scientific assumptions,
goals and values of functional researchers. He provides a succinct and accessible primer on
the functional level of analysis that is accessible to cognitive researchers. The philosophical
(functional contextualism) and theoretical (Relational Frame Theory) perspectives that have
guided research at this level are first unpacked and then a phenomenon known as arbitrarily
applicable relational responding (AARR) is discussed (also see Hughes et al., in press). For
nearly forty years now AARR has captured the imagination of functional researchers due to
its symbolic, flexible, and generative properties, with many arguing that it represents the basic
functional “building block” from which much of human psychological life springs forth. A
rising tide of studies seem to support this claim, with AARR linked to the origin and
INTRODUCTION TO SPECIAL ISSUE 19
development of many linguistic and cognitive abilities. The author showcases how an
analytic-abstractive perspective unlocks new insight into, and applications for, several areas in
psychological science, from language and rule-following, to analogical reasoning,
intelligence, theory of mind, psychopathology and implicit cognition.
Several contributions to the special issue consider concrete ways in which the FC
framework can be implemented. Liefooghe and De Houwer (this issue) provide a vivid
example of how the framework sets the stage for theoretical, methodological and empirical
developments within cognitive psychology. They focus their attention on cognitive control (a
collection of mental operations which enable humans to flexibly adapt in the face of changing
demands) and argue that a wide variety of procedures and effects have been used to study this
particular phenomenon. The authors then provide the blueprints for a functional approach to
cognitive control that has three main advantages. First, it leads to a general taxonomy of
cognitive control wherein effects are organized according to their functional commonalities
and differences. Second, it gives rise to a conceptual platform that facilitates communication
between cognitive researchers in this area. Third, it offers a framework for creating new tasks,
discovering new effects, and by implication, unlocking new questions about the processes that
underlie cognitive control.
Thereafter Perugini, Costantini, Hughes, and De Houwer (this issue) extend the
framework into the domain of personality psychology. They first draw attention to personality
research which typically focuses on how stable individual differences in behavior can be
carved into different categories at different levels of abstraction (e.g., traits and facets) with
the aim of predicting future outcomes. They then turn to the functional tradition which has
attained an impressive understanding of the behavioral principles underlying human behavior.
The authors consider how developments at the functional (analytic-abstractive) and mental
levels could be bridged in order to exploit the best of both worlds. Their functional-cognitive
INTRODUCTION TO SPECIAL ISSUE 20
account is both overarching in that it encompasses many if not all phenomena that are studied
in personality research and abstract in that it does so in terms of general behavioral principles
rather than superficial features of specific persons, situations, or behaviors. They conclude by
considering the potential of their approach for organizing existing scientific knowledge and
inspiring future research on personality.
De Houwer, Barnes-Holmes, and Barnes-Holmes (this issue) apply the framework
to clinical psychology with the aim of revealing differences and communalities among
Behavioral Therapy (BT), Cognitive Behavioral Therapy (CBT), and Acceptance and
Commitment Therapy (ACT). Drawing on the idea that functional and cognitive approaches
are situated at different levels of explanation, the authors argue that functional therapies such
as traditional BT and ACT are not necessarily incompatible with CBT and may actually
interact in a constructive manner. The degree to which these therapies align depends on
whether they point to the same or different types of environmental causes. The authors
propose that functional and cognitively oriented researchers and practitioners can therefore
engage in potentially fruitful interactions, while remaining true to their respective aims, so
long as the relationship between functional and cognitive explanations are explicated and the
two levels are firmly separated.
Whelan and Vahey (this issue) explore how the FC framework can be used to
enhance the study of the brain. They first shine a light on several problematic practices within
neuroscience with a particular emphasis on the treatment of behavioral/neural activity as
proxies for cognitive constructs and its various consequences. Rather that conflate the two
levels of explanation the authors call for those levels to be strictly delineated, with neural
activity first defined as behavior and the functional relation between that behavior and
environment identified. Thereafter mental operations can be deployed as heuristic tools to
explain what mediates between environmental input and neural output. Several illustrations of
INTRODUCTION TO SPECIAL ISSUE 21
how cognitive and functional neuroscientists can symbiotically interact to the benefit of both
communities are then offered.
In the third and final section, a number of scholars put forward their personal views on
the strengths and weaknesses of the framework. In their paper, Barnes-Holmes and Hussey
(this issue) argue that the functional (analytic-abstractive) approach is richly theoretical. They
first describe the steps involved in theory generation and evaluation at this level of analysis.
They then reflect upon the framework and arrive at three main conclusions. First, that
effective communication between the two traditions is likely to occur at the level of
behavioral observations rather than effects or theory. Second, not all behavioral observations
will be of mutual interest to both traditions. Third, observations that are of mutual interest will
be those that serve to elaborate and extend theorizing within a given tradition. The perceived
strengths and weaknesses of the framework are then discussed, along with the possibility that
it represents a third theoretical approach to psychological science rather than a meta-
theoretical perspective as initially proposed.
Proctor and Urcuioli (this issue) take a more critical stance and propose that certain
conditions will need to be addressed before effective communication between cognitive and
functional researchers can take place. They distinguish between two ways in which
“functional” can be referred to by the framework: as either referring to functional relations
that exist between environment and behavior (the effect-centric functional approach) or to an
specific intellectual tradition called Contextual Behavioral Science (CBS). They contend that
if the framework involves forging connections between CBS and cognitive science then it is
unlikely to meet with success given their contrasting scientific agendas. If the framework is
designed to ensure that behavioral effects are conceptually separated from the mental
operations used to explain those effects then this is a noble pursuit – but one that cognitive
researchers are already well acquainted with. For the authors the functional and cognitive
INTRODUCTION TO SPECIAL ISSUE 22
traditions are vying for scientific legitimacy and attempting to combine and even compare the
two may be a fool’s errand.
Finally, Fiedler (this issue) engages in a comparative evaluation of the functional and
cognitive levels of analysis. He argues that cognitive psychology is deeply anchored in
functional (effect centric) research and that progress within psychological science has often
been driven by accumulating knowledge of functional (environment-behavior) relations rather
than “stellar moments” of mental process research. The author attributes the functional level
primacy over the cognitive level for two reasons: (a) cognitive insight is contingent upon
environmental interventions and behavioral outcomes, and (b) knowledge about mental
operations and states needs to be cross-validated by returning to the functional level and
conducting further interventions and measurements. Nevertheless, he takes the view that the
functional and cognitive levels are deeply intertwined and jointly responsible for progress in
behavioral science.
Conclusion
We began our paper by comparing psychological science to an archipelago of
intellectual islands, each populated with a different tribe (sub-discipline) busily developing its
own culture (theories), language (terminology), tools (procedures) and (analytic) practices.
We argued that most of these islands subscribe to one of two belief systems (mental
mechanistic and behaviorism) that have historically propelled their proponents down two very
different scientific pathways. Many have, and continue to argue, that these differences are an
irreconcilable impediment to communication and collaboration, and that the two traditions
will be forever locked in combat for scientific legitimacy. The FC framework introduced here
takes a different stance. It sweeps such suggestions to the side and draws upon the knowledge
gained at the functional (effect-centric and analytic-abstractive) levels to advance theorizing
and research at the cognitive level (and vice-versa). Readers looking to test this claim need
INTRODUCTION TO SPECIAL ISSUE 23
only consider the strong selection of articles contained in this Special Issue. These showcase
how the FC framework is starting to (a) influence research and thinking in clinical, cognitive,
personality and neuropsychology in ways that (b) mutually benefit both cognitive and
functional researchers. Ultimately, only time will tell if the framework can achieve its
ambitious aims. But as travelers who have already journeyed to functional and cognitive
islands, and witness the value of their respective scientific fruits and labors, we believe that
this is one trip that others should not miss out on.
INTRODUCTION TO SPECIAL ISSUE 24
References
Anderson, N. H. (2013). Unified psychology based on three laws of information integration.
Review of General Psychology, 17(2), 125-132.
Barnes-Holmes, D., Barnes-Holmes, Y., Stewart, I., & Boles, S. (2011). A sketch of the
Implicit Relational Assessment Procedure (IRAP) and the Relational Elaboration and
Coherence (REC) model. The Psychological Record, 60(3), 527–542 .
Barnes-Holmes, D., & Hussey, I. (2015). The functional-cognitive meta-theoretical
framework: Reflections, possible clarifications and how to move forward.
International Journal of Psychology. Retrieved from
http://onlinelibrary.wiley.com/doi/10.1002/ijop.12166/full
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on cognitive
neuroscience. New York: Taylor & Francis.
Catania, A. C. (2007). Learning (Interim. 4a. Ed.) USA: Sloan Publishing.
De Houwer, J. (2011). Why the cognitive approach in psychology would profit from a
functional approach and vice versa. Perspectives on Psychological Science, 6(2), 202–
209.
De Houwer, J., Barnes-Holmes, D., & Moors, A. (2013). What is learning? On the nature and
merits of a functional definition of learning. Psychonomic Bulletin & Review, 20(4),
631–642.
De Houwer, J., Barnes-Holmes, Y., & Barnes-Holmes, D. (2015). Riding the waves: A
functional-cognitive perspective on the relations among behaviour therapy, cognitive
behaviour therapy and acceptance and commitment therapy. International Journal of
Psychology. Retrieved from http://onlinelibrary.wiley.com/doi/10.1002/ijop.12176/full
INTRODUCTION TO SPECIAL ISSUE 25
De Houwer, J., Gawronski, B., & Barnes-Holmes, D. (2013). A functional-cognitive
framework for attitude research. European Review of Social Psychology, 24(1), 252–
287.
Eagly, A. H., & Chaiken, S. (2007). The advantages of an inclusive definition of attitude.
Social Cognition, 25(5), 582–602.
Eysenck, M. W., & Keane, M. T. (2000). Cognitive psychology: A student’s handbook. New
York: Taylor & Francis.
Fazio, R. H. (2007). Attitudes as object-evaluation associations of varying strength. Social
Cognition, 25(5), 603-637.
Fiedler, K. (2014). From intrapsychic to ecological theories in social psychology: Outlines of
a functional theory approach. European Journal of Social Psychology, 44(7), 657–
670.
Gardner, H. (1985). The mind’s new science: A history of the cognitive revolution. New York:
Basic books.
Gawronski, B., & Bodenhausen, G. V. (2014). Theory and explanation in social psychology.
New York: Guilford Publications.
Gawronski, B., & Creighton, L. A. (2013). Dual process theories. In D. E. Carlston (Ed.), The
Oxford Handbook of Social Cognition (pp. 282–312). New York, NY: Oxford
University Press.
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(2), 175-
190.
Hughes, S., & Barnes-Holmes, D. (in press, a). Relational Frame Theory: An Overview. In
Handbook of Contextual Behavioral Science. New York: Wiley.
INTRODUCTION TO SPECIAL ISSUE 26
Hughes, S., & Barnes-Holmes, D. (in press, b). Relational Frame Theory: Implications for the
Study of Human Language and Cognition. In Handbook of Contextual Behavioral
Science (S. Hayes, D. Barnes-Holmes, R. Zettle, and T. Biglan (Eds.)). New York:
Wiley.
Hughes, S., De Houwer, J., & Barnes-Holmes, D. (in press). The Moderating Impact of Distal
Regularities on the Effect of Stimulus Pairings: A Novel Perspective on Evaluative
Conditioning. Experimental Psychology.
Hughes, S., De Houwer, J., & Perugini, M. (in press). Expanding the Boundaries of
Evaluative Learning Research: How Intersecting Regularities Shape Our Likes and
Dislikes. Journal of Experimental Psychology: General.
Krosnick, J. A., Judd, C. M., & Wittenbrink, B. (2005). The measurement of attitudes. The
Handbook of Attitudes, 21–76.
Liefooghe, B., & De Houwer, J. (2015). A functional approach for research on cognitive
control: Analyzing cognitive control tasks and their effects in terms of operant
conditioning. International Journal of Psychology. Retrieved from
http://www.researchgate.net/profile/Baptist_Liefooghe/publication/275885795_A_fun
ctional_approach_for_research_on_cognitive_control_Analyzing_cognitive_control_t
asks_and_their_effects_in_terms_of_operant_conditioning/links/554881020cf2e2031b
387b64.pdf
Miller, G. A. (2003). The cognitive revolution: a historical perspective. Trends in Cognitive
Sciences, 7(3), 141–144.
Nosek, B. A., Alter, G., Banks, G. C., Borsboom, D., Bowman, S. D., Breckler, S. J., …
others. (2015). Promoting an open research culture: author guidelines for journals
could help to promote transparency, openness, and reproducibility. Science (New York,
NY), 348(6242), 1422.
INTRODUCTION TO SPECIAL ISSUE 27
Perugini, M., Costantini, G., Hughes, S., & De Houwer, J. (2015). A functional perspective on
personality. International Journal of Psychology. Retrieved from
http://onlinelibrary.wiley.com/doi/10.1002/ijop.12175/full
Poldrack, R. A. (2006). Can cognitive processes be inferred from neuroimaging data? Trends
in Cognitive Sciences, 10(2), 59–63.
Roediger, H.L (2004). What happened to behaviorism. Observer, 17(3). Retrieved 21st
October 2015 from
http://www.psychologicalscience.org/index.php/uncategorized/what-happened-to-
behaviorism.html
Stewart, I. (2015). The fruits of a functional approach for psychological science. International
Journal of Psychology. Retrieved from
http://onlinelibrary.wiley.com/doi/10.1002/ijop.12184/full
Vahey, N., & Whelan, R. (2015). The functional-cognitive framework as a tool for
accelerating progress in cognitive neuroscience: On the benefits of bridging rather
than reducing levels of analyses. International Journal of Psychology. Retrieved from
http://onlinelibrary.wiley.com/doi/10.1002/ijop.12183/full
Zettle, R. D., Hayes, S. C., Biglan, T., & Barnes-Holmes, D. (Eds.). (in press). The handbook
of contextual behavioral science. Chichester, UK: Wiley/Blackwell.
