Understanding Metaphor: A Relational Frame Perspective

The Behavior Analyst 2001, 24, 191-199 No. 2 (Fall)

Understanding Metaphor:A Relational Frame Perspective

Ian Stewart and Dermot Barnes-HolmesNational University of Ireland, Maynooth

The current article presents a basic functional-analytic interpretation of metaphor. This work in-volves an extension of Skinner's (1957) interpretation of metaphor using relational frame theory(RFT). A basic RFI' interpretation of a particular metaphor is outlined, according to which themetaphor acquires its psychological effects when formal stimulus dimensions are contacted via thederivation of arbitrary stimulus relations. This interpretation sees the metaphor as involving fourelements: (a) establishing two separate equivalence relations, (b) deriving an equivalence relationbetween these relations, (c) discriminating a formal relation via this equivalence-equivalence rela-tion, and (d) a transformation of functions on the basis of the formal relation discriminated in thethird element. In the second half of the paper, a number of important issues with regard to the RFTinterpretation of metaphor are addressed.Key words: metaphor, relational frame theory, equivalence-equivalence, transformation of func-

tions, nonarbitrary relations

In psychology, especially cognitive psychology,characterizing the processes involved in thecomprehension of metaphor is not only an in-teresting challenge in its own right, but the spec-ification of those processes also constitutes agood test of the power of theories of languagecomprehension in general. (Ortony, 1993, p. 4)

Behavior analysis has had relativelylittle to say on the topic of metaphor.'Indeed, for the main contribution madein this area we have to turn to thelargely interpretive work of Skinner(1957). Skinner classified this phenom-enon as a subtype of the "extendedtact." The tact is a behavioral relationthat is defined as "a verbal operant inwhich a response of a given form isevoked (or at least strengthened) by aparticular object or event or property

Requests for reprints may be addressed toDermot Barnes-Holmes, Department of Psy-chology, National University of Ireland, May-nooth, Maynooth, County Kildare, Ireland (E-mail: [email protected]).

' Traditionally, a distinction is made betweenmetaphor and simile based largely on structuraldifferences (i.e., simile incorporates the phrases"like," "as," or "as if," whereas metaphor doesnot). Because the current interpretation andmodel of metaphor are functional analytic, thesestructural differences have been ignored for thetime being. Subsequent work may well discoverthat there is an important functional basis to thetraditional distinction between metaphor andsimile, but consideration of this issue is beyondthe scope of the current article.

of an object or event" (pp. 81-82). Theextended tact is a feature of more com-plex verbal behavior that occurs whena response is evoked by a novel stim-ulus that resembles a stimulus previ-ously present when a response was re-inforced. Metaphorical verbal behavioris a subtype of this latter class of be-havior that takes place "because of thecontrol exercised by properties of thestimulus which, though present at re-inforcement, do not enter into the con-tingency respected by the verbal com-munity" (p. 92). The following is anexample of Skinner's interpretation ofmetaphor:

When for the first time a speaker calls someonea mouse, we account for the response by notingcertain properties-smallness, timidity, silentmovement and so on-which are common to thekind of situation in which the response is char-acteristically reinforced and to the particular sit-uation in which the response is now emitted.Since these are not the properties used by zo-ologists or by the lay community as the usualbasis for reinforcing a response we call the ex-tension metaphorical. (p. 93)

A somewhat similar analysis of meta-phor, in which Skinner (1989) onceagain provided a behavior-analytic in-terpretation of familiar real-worldevents, appears in Recent Issues in theAnalysis of Behavior.

191

192 IAN STEWART & DERMOT BARNES-HOLMES

When we speak of weighing evidence we areusing a metaphor. But a metaphor is a word thatis "carried over" from one referent to anotheron the basis of a common property. The com-mon property in weighing is the conversion ofone kind of thing (potatoes or evidence) into an-other (a number on a scale or a verdict). Oncewe have seen this kind of thing done with po-tatoes it is easier to see it done with evidence.... We could also say that weight becomes ab-stract when we move from potatoes to evidence.The word is indeed abstracted in the sense of itsbeing drawn away from its original referent, butit continues to refer to a common property, and,as in the case of metaphor, in a possibly moredecisive way. The testimony in a trial is muchmore complex than a sack of potatoes, and"guilty" probably implies more than "tenpounds." But abstraction is not a matter of com-plexity. Quite the contrary. Weight is only oneaspect of a potato, and guilt is only one aspectof a person. Weight is as abstract as guilt. It isonly under verbal contingencies of reinforce-ment that we respond to single properties ofthings or persons. In doing so we abstract theproperty from the thing or person. (p. 7)

According to the Skinnerian account,therefore, metaphor may be conceptu-alized as the abstraction, via one par-ticular subtype of verbal behavior (i.e.,the extended tact), of a common phys-ical property from two different typesof environmental event. This behavior-al interpretation, however, does not de-tail how a repertoire of metaphoricalverbal behavior (especially complexhuman metaphor) develops from a pre-sumably simpler repertoire of formalproperty abstraction. This is perhapsthe key question left unanswered bySkinner's analysis. However, Skinner'swork does provide the basis for a mod-em behavioral treatment of metaphorusing the conceptual and empiricalframework of relational frame theory(RFT; Hayes, Barnes-Holmes, &Roche, 2001; see also Barnes-Holmes,Barnes-Holmes, & Cullinan, 2000, fora systematic synthesis of the Skinner-ian and RFT interpretations of verbalbehavior).

RELATIONAL FRAME THEORY

Relational frame theory is an ap-proach to human language and cogni-tion that treats these phenomena as ex-amples of arbitrarily applicable rela-

tional responding. All animals that arecapable of complex forms of learningmay be trained to make discriminationson the basis of nonarbitrary or physicalrelations between stimuli (e.g., louderthan, larger than, differently coloredfrom). However, according to RFT,language-able humans, having pro-longed exposure to certain contingen-cies of reinforcement that operate with-in the verbal community, also demon-strate responding on the basis of de-rived or arbitrarily applicable relations.These relations are defined not by thephysical properties of the relata per sebut by additional contextual cues.One important example of arbitrarily

applicable relational responding is thephenomenon of stimulus equivalence.For example, if a language-able humanis taught to match an arbitrary Stimu-lus A to an arbitrary Stimulus B and tomatch Stimulus B to an arbitrary Stim-ulus C, then, without further training,he or she might subsequently match Ato B, B to C (thus showing symmetri-cal responding), C to A (thus showingtransitive responding), and A to C(thus showing combined symmetricaland transitive responding). In otherwords, the person now treats the threestimuli as mutually substitutable orequivalent.

Relational frame theory takes theview that stimulus equivalence, andwhat has been called a transfer of func-tion through equivalence relations, pro-vide an important basis for the behav-ior analysis of symbolic relations andhuman language more generally. Froman RFT perspective, one of the impor-tant properties of a symbol is that itspsychological functions are based tosome extent on the transfer of func-tions through equivalence relations.Suppose, for example, that a youngboy enjoys a glass of Cokes (StimulusA) for the first time and is then told bya school friend that Coke is like anoth-er drink called PepsiT3 (Stimulus B). Ifthe boy is subsequently asked "Wouldyou like a Pepsi?" he may well re-spond positively. This transfer of ap-petitive function is based on the di-

UNDERSTANDING METAPHOR 193

rectly established psychological func-tion of the word Coke, which waspaired with actual Coke, and the de-rived symmetry relation between thewords Pepsi and Coke (i.e., Coke islike Pepsi). In effect, the boy does notneed to experience Pepsi directly in or-der for him to "think" that it will be apleasant drink. This basic concept of atransfer of function in accordance withstimulus relations forms the core of theRIFT approach to language and cogni-tion (Hayes, Barnes-Holmes, & Roche,2001).

In this article we will describe howRFT has used the concepts of equiva-lence and transfer of function to ana-lyze metaphor both conceptually andempirically. This RFT interpretationbuilds on the earlier work of Skinner(1957) by showing how a repertoire ofmetaphorical verbal behavior mightdevelop from a simpler repertoire offormal property abstraction. Beforecontinuing however, we must first in-clude a caveat.

The model reported herein adds toSkinner's (1957) analysis of metaphor-ical extension by showing that it canbe based on arbitrarily applicable re-lations among stimuli. Of course, in1957, Skinner did not have access tothe concepts of equivalence and trans-fer of function that are necessary to de-scribe the conditions from which arbi-trary stimulus relations emerge. Be-havior-analytic science can now de-scribe some of these conditions.However, to claim that we have isolat-ed them sufficiently goes beyond cur-rently available data. Under a varietyof conditions, with a variety of tests,investigators have revealed the limita-tions of our understanding of equiva-lence and derived relational respondingmore generally, and thus at this time,we are left with empirical questions yetto be answered with regard to thesephenomena (e.g., Fields, Hobbie-Reeve, Adams, & Reeve, 1999; Pil-grim & Galizio, 1995; Spencer &Chase, 1996). It should be recognized,therefore, that there are limits to theRFT interpretation of metaphor report-

ed below that will be overcome onlythrough further empirical research.

RELATING RELATIONS:A KEY PROPERTY OF

METAPHORICAL LANGUAGE

The first published RFT study thatwas directly relevant to the analysis ofmetaphor focused on the closely relat-ed area of analogy (Barnes, Hegarty, &Smeets, 1997; we will consider the re-lation between analogy and metaphorat a later point in the article). Specifi-cally, Barnes et al. used what theycalled equivalence-equivalence re-sponding to construct a relationalframe interpretation of analogical rea-soning. In the authors' own words,

Consider ... the following question based on theclassic proportion scheme (A: B :: C: ?); "ap-ple is to orange as dog is to (i) sheep, or (ii)book?" If "apple" and "orange" participate inan equivalence relation in the context "fruit,"and "dog" and "sheep" participate in an equiv-alence relation in the context "animals," thenwe would expect a person to pick "sheep" asthe correct answer. In effect, the response wouldbe in accordance with the derived equivalencerelation between two already established sepa-rate equivalence relations. ... We take the viewthat equivalence-equivalence responding is anexample of a relational network as defined byrelational frame theory (e.g., Barnes, 1994;Barnes & Holmes, 1991; Hayes, 1991, 1994).(1997, p. 3)

The first experiment reported byBarnes et al. (1997) examined the re-lations between two separate equiva-lence relations and between two sepa-rate nonequivalence relations. Subjectswere first trained and tested for the for-mation of four three-member equiva-lence relations (i.e., train Al -e B1, Al-+ Cl,A2 B2,A2 e> C2,A3 -+ B3,A3 - C3, A4 - B4, A4 - C4, andtest Bi <-+ Cl, B2 e-> C2, B3 <-> C3,B4 e- C4). After successfully passingthe equivalence test, subjects were test-ed to determine whether they would re-late pairs of stimuli to other pairs ofstimuli based on their participation inequivalence relations. Subjects werepresented with samples that containedtwo stimuli that were from one derivedequivalence relation (e.g., BICI), and


were given the opportunity to choosecomparisons that contained two stimulithat were from a second separate de-rived equivalence relation (e.g.,B3C3). The result of this first experi-ment was that a range of subjects, in-cluding a 12-year-old boy, successfullyrelated equivalence relations to otherseparate equivalence relations, andnonequivalence relations (e.g., B1B2)to other, separate nonequivalence rela-tions (e.g., C3C4), in the absence ofexplicit reinforcement. Experiment 2employed the same procedures as wereused in Experiment 1, except that sub-jects were exposed to the equivalence-equivalence test before being exposedimmediately, and without further train-ing, to the standard equivalence test.Again, the result of this experimentwas that all subjects, this time includ-ing a 9-year-old boy, successfully re-lated equivalence relations to otherseparate equivalence relations and non-equivalence relations to other separatenonequivalence relations. In short,Barnes et al. provided an empiricalmodel of analogical reasoning, basedon the RFT concept of equivalence-equivalence responding. This basicconcept of relating one derived relationto another derived relation also lies atthe heart of the RFT analysis of met-aphor, to which we now turn.

METAPHOR FROMTHE RFT PERSPECTIVE

Consider the metaphor, "Strugglingwith anxiety is like struggling in quick-sand," which might be used, for ex-ample, in the psychotherapeutic treat-ment of clinical anxiety. Contactingthis particular metaphor might cause aclient to change his or her response toanxiety attacks. The client probably al-ready knows, via the verbal commu-nity, that struggling in quicksand onlymakes drowning all the more likely.When the therapist suggests that strug-gling with anxiety is similar to strug-gling in quicksand, then the client maysee that struggling with anxiety servesonly to make the anxiety worse (i.e.,

trying really hard to escape feelings ofanxiety can often increase those veryfeelings). Consequently, the client's be-havior in the face of anxiety maychange, such that he or she no longerattempts strongly to resist his or heranxiety, but instead accepts the anxiousfeelings as they arise, and thus pre-vents the downward cycle into full-blown panic.

This quicksand-anxiety metaphormay be interpreted as a relational net-work that is functionally similar to theanalogy described earlier. In this case,however, "anxiety" (A) is to "psycho-logical struggle" (B) as "quicksand"(C) is to "physical struggle" (D) (seeFigure 1, Elements 1 and 2). For cur-rent purposes, we will describe the re-lation between anxiety and psycholog-ical struggle, and the relation betweenquicksand and physical struggle, astwo separate equivalence relations.From the RFT perspective, the rela-tional network in this example of met-aphor may help the listener to discrim-inate aformal relation between two ap-parently very different events. Thisdiscrimination of formal similarity be-tween the two events may make atransfer of function from one to theother more likely. In this particularmetaphor, deriving a relation betweenanxiety and psychological struggle andquicksand and physical struggle couldhelp the listener to discriminate thatstruggling in either case leads to struc-turally or formally similar physiologi-cal and psychological effects (i.e., mas-sively increased autonomic arousal anda sense of fear and panic; see Figure 1,Element 3). Consequently, some of thefunctions of "quicksand" might nowbe more likely to transfer to "anxiety."For example, a clinically anxious per-son might derive certain importantcause-effect relations (these and othernonequivalence relations will be con-sidered below) including the following:"Struggling with anxiety will onlymake my situation worse" and "byceasing to struggle I can begin to over-come my anxiety" (see Figure 1, Ele-ment 4). Thus, the original problem


Element (i): In the dients original relatonal network, "quicksand" has certainimportant response functions such as "Dont struggle", for example, which"anxiety" does not.

ANXIETY QUICKSAND (fn.: Don't struggle)

Equiv Equiv

PSYCHOLOGICAL STRUGGLE PHYSICAL STRUGGLE

Element (ii): The therapist changes the dient's relational network by equatingthe relation "anxiety [AJ / psychological struggle [Br with the relation"quicksand [C] / physical struggle [DI " (i.e., "A is to B as C is to D").

[A] ANXIETY QUICKSAND (fn.) [C]

Equiv t- EQUIV-> Equiv

[B] PSYCHOLOGICAL STRUGGLE PHYSICAL STRUGGLE [D]

Element (iii): The newly changed relabonal network allows the client todiscriminate a formal or non arbitrary relation of sameness between therelations "anxiety / struggle" and "quicksand / struggle" (i.e., both relationslead to increasing autonomic arousal accompanied by spiraling feelings offear and panic).

ANXIETY QUICKSAND (fn.)

Equiv <- EQUIV-> Equiv


(Increasing autonomic (Increasing autonomicarousal / Fear) arousal / Fear)

I(FORMAL/NON-ARBITRARY RELATION)

Element (iv): The discrimination of a formal or non arbitrary relation betweenthe relations "anxiety / struggle" and "quicksand / struggle" may lead to atransfer of the funcfions of "anxiety" such that "anxiety" now possesses some ofthe response functions of quicksand including, for example,"Don't struggle";

ANXIETY (fn.: Don't struggle) QUICKSAND (fn.: Don't struggle)

Equiv <- EQUIV-> Equiv


I(Increasing autonomic === (Increasing autonomic

arousal / Fear) arousal / Fear)I

(FORMAL/NON-ARBITRARY RELATION)

Figure 1. The four elements that, from the RFT perspective, characterize the process of metaphor.

may come to be "recast" in view ofthe relational network illustrated inFigure 1 (Elements 2 through 4). Moregenerally, this type of pattern of trans-fer of functions throughout a relationalnetwork generates, from the RFT per-

spective, the often-experienced rich-ness and complexity of metaphoricallanguage, as well as the emotional "in-sight" that metaphor can often conferupon the listener. Parenthetically, weshould emphasize that the four ele-


ments involved in metaphorical re-sponding are not stages, and that theexact sequence of these and perhapsother elements may vary from instanceto instance. We will return to this issuelater in the article.

THE RFT INTERPRETATIONOF METAPHOR:

ADDITIONAL ISSUESThe foregoing provides an RFI in-

terpretation of how metaphor may beused to change the behavior of a lan-guage-able human in important ways.In fact, we have developed a four-element empirical model based onthis RFT conceptualization (Stewart,Barnes-Holmes, & Roche, 2000). Nev-ertheless, our conceptual and empiricalwork constitutes only a beginning, andthere are many issues that remain to beaddressed.

Metaphor and Relations OtherThan Equivalence

According to RFT, equivalence isjust one example of derived or arbi-trarily applicable relational responding.Relational frames of opposite, differ-ence, before-after, and so forth havebeen the subject of both conceptualand empirical investigation, therebyextending the range of behavioral phe-nomena that might emerge fromtrained relational responding (e.g.,Barnes & Hampson, 1993a, 1993b;Barnes & Keenan, 1993; Dymond &Barnes, 1994; Roche & Barnes, 1997;Steele & Hayes, 1991).

In the current article, we have pre-sented an RFT interpretation of meta-phor based on equivalence-equiva-lence responding, that is, one in whichthe listener responds in accordancewith a relation of equivalence betweentwo equivalence relations. However,according to RFT, metaphors also mayinvolve relations of equivalence be-tween other types of relation. For ex-ample, the metaphor "Struggling withanxiety is like struggling in quick-sand" might be more accurately de-scribed as an equivalence relation be-

tween two causal relations (i.e., a"causal-causal" relation). Specifically,struggling causes more rapid sinkinginto either quicksand or anxiety. Fur-ther research in our laboratories is cur-rently utilizing novel methodologies(e.g., Hayes & Barnes, 1997) to modelmetaphorical language in which non-equivalence relations participate in theunderlying relational networks.

The Distinction Between Metaphorand Analogy

The current interpretation of meta-phor proposes that the key propertiesof metaphorical language include, forexample, the relating of relations andthe discrimination of nonarbitrary sim-ilarity via arbitrarily applicable rela-tional responding. However, theseproperties could also be seen as char-acterizing analogy. Consider, for ex-ample, the classic Rutherfordian anal-ogy "An atom is like the solar sys-tem." This analogy involves the relat-ing of relations. That is, in both cases,there is a relation between a centralhub and orbiting elements, and becausethese are similar relations they may betreated as equivalent. Nonarbitrary re-lations are also involved, because therelation between the central hub andorbiting elements in both systems is anobviously physical or nonarbitrary re-lation. Given that analogy involves re-lating relations and nonarbitrary rela-tions, the question remains: How doesmetaphor differ from analogy, and howmight we interpret this difference infunctional terms?One test that has been used to dis-

tinguish between analogy and meta-phor is the directionality test (see, e.g.,Chase, 1986), according to which anal-ogies are bidirectional, whereas meta-phors are unidirectional. In any exam-ple of either analogy or metaphor, twoevents, A and B, are related in a phraseof the basic form "A is (like) B." Inthe case of analogy, the position of theA and B terms may be swapped andthe result is another meaningful anal-ogy. For example, if the A and B terms


in the analogy "An atom (A) is like thesolar system (B)" are swapped, the re-sult, "The solar system (B) is like anatom (A)" is still a valid and under-standable analogy. However, in thecase of metaphor, if the A and B termsare swapped, the phrase loses its met-aphorical quality. Consider the meta-phor "Cats (A) are dictators (B)," forexample. Metaphors such as this workbecause the A and B terms have aproperty in common that is obviousand stereotypical in the case of B butis not obvious in the case of A. In thisexample, dictators are clearly demand-ing and coercive, whereas cats mayalso be seen in this light but far lessobviously. The inclusion of these termsin the metaphor in the order cats (A)are dictators (B) serves to amplify theproperty of "demanding-ness" in theA term, "cats." However, this ampli-fication effect occurs only if this spe-cific order is maintained. If the termsare swapped, so as to yield the phrase,"Dictators are cats," then there is noappropriate amplification effect, be-cause now there is no longer a com-mon property that is obvious in thesecond term and not obvious in thefirst.The earlier example, "Struggling

with anxiety is like struggling in quick-sand," also qualifies as a metaphorbased on this definition, because theproperty in common between the twoevents (i.e., self-defeating struggle) isobvious in the second term but not inthe first. Hence, the juxtapositionworks in the direction just shown butnot so well in the other direction.A technical RFT account of the dif-

ferent directionalities characterizinganalogy and metaphor is as follows. Inthe case of analogies, such as "Anatom (A) is like the solar system (B),"the nonarbitrary properties that A andB have in common are equally char-acteristic of both A and B. Thus, therewill be a relation of equivalence be-tween the A and B terms, and reversalof the order of the terms still yields ananalogy. In the case of metaphor, thereare often contextual cues for the deri-

vation of a hierarchical frame betweenthe two elements involved (see Hayes,Fox, et al., 2001). For example, En-glish speakers respond to a sentence ofthe form "A is B" (e.g., apples arefruit) as signifying that A are membersof the class of B, and thus that A (e.g.,apples) have at least some of the typi-cal properties of B (e.g., fruit). That is,"apple" is contained in the class"fruit," but "fruit" is not contained inthe class "apple." This hierarchicalframe is an important property of met-aphor. In technical terms, "Cats aredictators" may cause a transfer of thepsychological functions typically as-sociated with dictators (e.g., "demand-ing-ness") to cats, and if some of thenonobvious functions of cats are am-plified by this transfer, the listener willrespond to it as a metaphor. However,if the A and B terms are reversed, theobvious qualities of cats transfer to dic-tators and amplification is less likely(e.g., dictators have none of the quali-ties of small, furry pets). In effect, theunidirectional, hierarchical relation be-tween the terms involved is an impor-tant feature of the phenomenon of met-aphor, because it differentiates it fromanalogy.

Elements, Not Stages

As pointed out earlier, the four-ele-ment interpretation of metaphor is notsequential-the exact sequence of ele-ments may vary from instance to in-stance. The formal relations specifiedin Element 3, for example, could serveas a contextual cue for equivalence re-sponding or could itself be discrimi-nated only after the equivalence rela-tion is derived. Consider again the met-aphor, "Cats are dictators." As waspointed out, the shared formal featuresof the two terms are their demandingqualities. This common quality may it-self function as a context for applyinga frame of equivalence between catsand dictators (i.e., Element 3 may oc-cur before, or at least feed back to sup-port, Element 2). Furthermore, certainelements (e.g., Element 4) may be


missing entirely in some metaphors(i.e., the metaphor is understood, butthere is no change in behavior). Nev-ertheless, the four elements presentedin this example are typical of many in-stances of metaphor.

Understanding Metaphor VersusCreating Metaphor

The current work has focused on theunderstanding of metaphorical lan-guage as opposed to the creation ofnovel metaphors. Future research,however, will need to focus on the un-doubtedly important behavioral pro-cesses involved in the creation of novelmetaphors. Within the context of thecurrent interpretation of understandingmetaphor, the creator of a metaphormight first perceive some formal ornonarbitrary similarity between twodifferent environmental events, andthen subsequently behave verbally inaccordance with a new relational net-work into which this novel formal sim-ilarity has been incorporated. The cre-ator of the metaphor "surfing the In-ternet," for example, might have beenusing the Internet at some point andfound that this context suddenly pro-duced some of the perceptual functionsof actual surfing. Consequently, he orshe may have immediately related thetwo terms, "surfing waves" and "surf-ing the Internet," thus creating a rela-tional network on the basis of nonar-bitrary relations.

Although we have argued that thecreation of novel metaphorical lan-guage may involve the discriminationof formal relations prior to the firstdemonstration of the new metaphor, wewould also argue that a prior history ofarbitrary relational responding is nec-essary for the initial discrimination ofthe nonarbitrary relations to functionas the basis for metaphor (i.e., verbalhistory -> nonarbitrary relations -o ar-bitrary relations). The metaphor, "surf-ing the Internet," for example, is sure-ly based, in part, on the verbal historythat gave rise to descriptions of bothsurfing the sea (e.g., "surfing is fast

and exciting and feels like 'skimming'or 'riding' the sea") and using the In-ternet (e.g., "accessing information onthe Internet is fast, and it feels as if theuser is 'skimming' or 'riding' over asea of information"). In other words,these descriptions facilitated the dis-crimination of the formal relations thatled to the production of the metaphorfor the first time. From this perspec-tive, therefore, the creation of meta-phor involves the discrimination ofnonarbitrary relations based on a his-tory of arbitrary relational responding,and thus the processes involved in un-derstanding and creating metaphormay be functionally similar. Althoughthis issue will certainly require furtheranalysis, the current RFT model of un-derstanding metaphor should provide auseful starting point for developing amore complete functional-analyticmodel of metaphorical language.

REFERENCES

Barnes, D. (1994). Stimulus equivalence andrelational frame theory. The PsychologicalRecord, 44, 91-124.

Barnes, D., & Hampson, P. J. (1993a). Learningto learn: The contribution of behavior analysisto connectionist models of inferential skill inhumans. In G. Orchard (Ed.), Neural comput-ing research and applications (Vol. 1, pp.129-138). London: IOP.

Barnes, D., & Hampson, P. J. (1993b). Stimulusequivalence and connectionism: Implicationsfor behavior analysis and cognitive science.The Psychological Record, 43, 617-638.

Barnes, D., Hegarty, N., & Smeets, P. M.(1997). Relating equivalence relations toequivalence relations: A relational framingmodel of complex human functioning. TheAnalysis of Verbal Behavior, 14, 57-83.

Barnes, D., & Holmes, Y (1991). Radical be-haviorism, stimulus equivalence, and humancognition. The Psychological Record, 41, 19-31.

Barnes, D., & Keenan, M. (1993). A transfer offunctions through derived arbitrary and non-arbitrary stimulus relations. Journal of the Ex-perimental Analysis of Behavior, 59, 61-81.

Barnes-Holmes, D., Barnes-Holmes, Y, & Cul-linan, V. (2000). Relational frame theory andSkinner's Verbal Behavior: A possible synthe-sis. The Behavior Analyst, 23, 69-84.

Chase, P N. (1986). Three perspectives on ver-bal learning: Associative, cognitive and op-erant. In P N. Chase & L. J. Parrott (Eds.),Psychological aspects of language: The West


Virginia lectures (pp. 5-35). Springfield, IL:Charles C Thomas.

Dymond, S., & Barnes, B. (1994). A transferof self-discrimination response functionsthrough equivalence relations. Journal of theExperimental Analysis of Behavior, 62, 251-267.

Fields, L., Hobbie-Reeve, S. A., Adams, B. J.,& Reeve, K. J. (1999). Effects of training di-rectionality and class size on equivalence classformation by adults. The Psychological Rec-ord, 49, 703-724.

Hayes, S. C. (1991). A relational control theoryof stimulus equivalence. In L. J. Hayes & PN. Chase (Eds.), Dialogues on verbal behav-ior: the first international institute on verbalrelations (pp. 19-40). Reno, NV: ContextPress.

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 andcognition (pp. 9-30). Reno, NV: ContextPress.

Hayes, S. C., & Barnes, D. (1997). Analyzingderived stimulus relations requires more thanthe concept of stimulus class. Journal of theExperimental Analysis of Behavior, 68, 235-270.

Hayes, S. C., Bames-Holmes, D., & Roche, B.(2001). Relationalframe theory: A post-Skin-nerian approach to language and cognition.New York: Plenum.

Hayes, S. C., Fox, E., Gifford, E. V., Wilson, K.G., Barnes-Holmes, D., & Healy, 0. (2001).

Derived relational responding as learned be-havior. In S. C. Hayes, D. Barnes-Holmes, &B. Roche (Eds.) Relational frame theory: Apost-Skinnerian approach to language andcognition (pp. 21-49). New York: Plenum.

Ortony, A. (1993). Metaphor and thought.Cambridge, England: Cambridge UniversityPress.

Pilgrim, C., & Galizio, M. (1995). Reversal ofbaseline relations and stimulus equivalence: I.Adults. Journal of the Experimental Analysisof Behavior, 63, 225-238.

Roche, B., & Barnes, D. (1997). A transfor-mation of respondently conditioned stimulusfunction in accordance with arbitrarily appli-cable relations. Journal of the ExperimentalAnalysis of Behavior, 67, 275-301.

Skinner, B. F (1957). Verbal behavior. NewYork: Appleton-Century-Crofts.

Skinner, B. F (1989). Recent issues in the anal-ysis of behavior. Columbus, OH: Merrill.

Spencer, T. J., & Chase, P. N. (1996). Speedanalyses of stimulus equivalence. Journal ofthe Experimental Analysis of Behavior, 65,643-659.

Steele, D., & Hayes, S. C. (1991). Stimulusequivalence and arbitrarily applicable rela-tional responding. Journal of the Experimen-tal Analysis of Behavior, 56, 519-555.

Stewart, I., Barnes-Holmes, D., & Roche, B.(2000, May). Advancing the experimentalanalysis of metaphor with relational frametheory. Paper presented at the annual conven-tion of the Association for Behavior Analysis,Washington, DC.

Documents

Understanding Metaphor: A Relational Frame Perspective