I was recently reading about the Relational Frame Theory of Hayes, Barnes-Holmes, and others, and I´d like to know your thoughts about this research program.

katja z wrote:Hmm, I'm not really familiar with this, but I gather that it is essentially about language acquisition through operant conditioning, is that right? This seems reasonable, and far more promising than the various incarnations of language acquisition device.

Can you tell some more and/or link to any interesting articles?

Yes, it´s an operant theory of language (not only about its acquisition, but also about its characteristics and causal relations). Here are some links:
Wikipedia: http://en.wikipedia.org/wiki/Relational_frame_theory
Book: http://ifile.it/ej583yn/0306466007.rar
Tutorial (in Flash): http://www.4shared.com/account/video/0Opf0-8M/Tutorial_RFT__Eric_Fox_.html
Articles: http://contextualpsychology.org/suggested_rft_readings

Here´s a brief summary:
There´s a phenomenon called “stimulus equivalence”, in which after a training phase of conditional discriminations (given the sample S1, selection of S2 is reinforced, given the sample S2, selection of S3 is reinforced), the animal shows untrained derived behaviors of simmetry (given S2, select S1), transitivity (given S1, select S3) and function transfer (if one stimulus changes its function, the other stimuli also change in the same way).
Why does this phenomenon occur, if those derived behaviors are not trained in the experiment? Relational Frame Theory (RFT) proposes that they´re learned before the experiment, through reinforcement across multiple exemplars. Reinforcement is not the reductionistic version imagined by Chomsky (an explicit praise or a candy): it can be thought of as the many environmental causes of long term potentiations of the neural components of a particular response. The child learns to say a name after seeing a referent, and to look for a referent after hearing a name. After multiple exemplar training, the child learns to show the derived simmetrical response (given A then B, given B then A) in many situations. This is an operant conditioning, where the antecedent become a discriminative stimulus for the operant behavior of showing derived simmetrical responses. The same kind of training occurs for transitivity and function transfer
But children also learn other relational responses that are more complex than equivalence. They learn, for example, that the relation “is bigger than” is transitive but not simmetrical. So each kind of relation is learned as an operant defined by 3 components: mutual entailment (of which simmetry is only an example), combinatory entailment (of which transitivity is only an example), and function transformations (of which function transfer is only an example). The behavior of deriving relations (mutual and combinatory entailment) and transforming functions is called "relational operant", and is controlled by antecedent discriminative stimuli. Each kind of relational operant is called a "relational frame", for example: “A is simmilar to B” (frame of coordination), “A is opposite to B” (frame of opposition), “A is different from B” (frame of distinction), “A is bigger than B” (frame of comparison), “A is better than B” (frame of comparison), “A is part of B” (hierarchical frame), “A occurs after B” (temporal frame), “A is condition for B” (conditional frame), “A is mine, B is yours” (deictic frame), etc.
I´ll continue later...

I don't know a whole lot about this subject, but I do know that whenever I've heard behavioral scientists (particularly those working in stimulus equivalence research) discuss relational frame theory, it's always in a negative way. Some critics include:

Laudable goals, interesting experiments, unintelligible theorizing: A critical review of Relational Frame Theory - Burgos

Data In Search Of A Principle: A Review Of Relational Frame Theory: A Post-Skinnerian Count Of Human Language And Cognition - Palmer

Function transfer in human operant experiments: the role of stimulus pairings - Tonnaeu and González

And the best critique is probably:

Who Can Understand Relational Frame Theory? A Reply to Barnes-Holmes and Hayes - Tonneau

One of the main issues with the theory put forward by researchers is that its mechanisms are so vague that it makes it difficult to tell exactly what they're proposing, and how we could possibly test certain aspects of it. Perhaps more damning, however, is the issue discussed by Tonneau in that last paper (under the chapters titled "Associations and Reinforcement" and "Associations and RFT") where the proponents of RFT appear to completely discount the role of Pavlovian conditioning in their explanations of how stimulus equivalence operates. This is obviously a big problem since we know that Pavlovian conditioning is a massive component of stimulus equivalence, arguably more so than operant conditioning.

Generally, RFT is ignored in the literature and the main theories for the explanation of stimulus equivalence (and thus language) are "Sidman's 2000 theory" (highly imaginative name there) and Tonnaeu's "Stimulus-Relation" theory. More recently, people have rejected Sidman's theory as being incomplete and failing to adequately account for the importance of Pavlovian conditioning, and so the research has been pushed towards Tonneau's theory as it appears to be the best account of stimulus equivalence. This paper demonstrates why Sidman's account is inferior: Stimulus Equivalence: Testing Sidman's (2000) Theory - Minster, Elliffe and Muthukumaraswamy.

Mr.Samsa wrote:I don't know a whole lot about this subject, but I do know that whenever I've heard behavioral scientists (particularly those working in stimulus equivalence research) discuss relational frame theory, it's always in a negative way. Some critics include:
Laudable goals, interesting experiments, unintelligible theorizing: A critical review of Relational Frame Theory - Burgos
Data In Search Of A Principle: A Review Of Relational Frame Theory: A Post-Skinnerian Count Of Human Language And Cognition - Palmer
Function transfer in human operant experiments: the role of stimulus pairings - Tonnaeu and González
And the best critique is probably:
Who Can Understand Relational Frame Theory? A Reply to Barnes-Holmes and Hayes - Tonneau

There´re replies to each of those critics: Hayes et al. (2003) have answered Burgos (2003), Hayes & Barnes-Holmes (2004) have answered Palmer (2004), and Hayes (2002) has answered Tonneau (2002). For an accurate assessment, we need to compare the respective arguments and counter-arguments in detail. My own conclusion is that RFT is not a panacea, and it needs to be improved in many of its characteristics, but it´s still a plausible and progressive research program, best than most of its actual rivals.
There´s a review of this controversy in Gross and Fox (2009):
http://contextualpsychology.org/system/files/GrossFox_RFTControversy.pdf

Mr.Samsa wrote:One of the main issues with the theory put forward by researchers is that its mechanisms are so vague that it makes it difficult to tell exactly what they're proposing, and how we could possibly test certain aspects of it. Perhaps more damning, however, is the issue discussed by Tonneau in that last paper (under the chapters titled "Associations and Reinforcement" and "Associations and RFT") where the proponents of RFT appear to completely discount the role of Pavlovian conditioning in their explanations of how stimulus equivalence operates. This is obviously a big problem since we know that Pavlovian conditioning is a massive component of stimulus equivalence, arguably more so than operant conditioning.

How would you justify your assessment that “pavlovian conditioning is arguably more important than operant conditioning”? I agree that pavlovian conditioning seems to be a relevant part of the phenomenon, but that´s not enough to say that it´s “more important” than operant conditioning, or that an operant account such as RFT can be dismissed. Also, there´re a lot of empirical and theoretical controversies about the differences and relations between pavlovian and operant conditioning (including some proposals that they´re different experimental conditions with simmilar underlying neural mechanisms), and those controversies should be discused and clarified before arguing such conclusion.

Mr.Samsa wrote:Generally, RFT is ignored in the literature and the main theories for the explanation of stimulus equivalence (and thus language) are "Sidman's 2000 theory" (highly imaginative name there) and Tonnaeu's "Stimulus-Relation" theory.

I don´t think so, and we could easily test this claim with a bibliometric assessment of some journals. RFT is not “ignored” in the literature of stimulus equivalence and language. On the contrary, it´s considered one of the most relevant rival accounts (even by its critics, as the ones that you´ve mentioned). The most mentioned accounts of stimulus equivalence and language in the behavior-analytic literature are: RFT, Sidman´s theory, Tonneau´s theory, and Horne & Lowe´s “naming” theory. Also, unlike the other research programs that you´ve mentioned, RFT researchers are studying many of the more relevant aspects of language (they are doing research about rules, instructions, analogies, metaphores, mental concepts, moral concepts, emotional concepts, reasoning, grammar). None of the other research programs are doing studies of so many aspects of language. See also the interesting convergencies between RFT and cognitive research:
http://www.nuigalway.ie/psychology/documents/dale_2002.pdf
http://cognaction.org/rick/pdfs/papers/dale_2004.pdf

seeker wrote:There´re replies to each of those critics: Hayes et al. (2003) have answered Burgos (2003), Hayes & Barnes-Holmes (2004) have answered Palmer (2004), and Hayes (2002) has answered Tonneau (2002). For an accurate assessment, we need to compare the respective arguments and counter-arguments in detail. My own conclusion is that RFT is not a panacea, and it needs to be improved in many of its characteristics, but it´s still a plausible and progressive research program, best than most of its actual rivals.

Well what is it about RFT that you find so convincing? I've heard of people who find the concept plausible, but I don't know any who seem quite as convinced as you (except perhaps Hayes and Barnes-Holmes, or therapists).

seeker wrote:There´s a review of this controversy in Gross and Fox (2009):
http://contextualpsychology.org/system/files/GrossFox_RFTControversy.pdf

To be honest, this paper just confused me further. This part for example:

According to RFT, arbitrarily applicable
relational responding is the foundation of
human language and cognition; hence, the
definition of verbal behavior is simply ‘‘the
action of framing events relationally’’
(Hayes, Fox, et al., 2001, p. 43). Accordingly,
the definition of verbal stimuli is ‘‘stimuli
that have their effects because they participate
in relational frames’’ (Hayes, Fox, et al.,
2001, p. 44). The history of the acting
organism is the basis for bringing about
verbal stimulus functions, not the history of
another organism or listener. In the RFT
analysis, both the speaker and the listener are
engaging in verbal behavior. The speaker
does so by producing stimuli that are based
on relationally framed events, and the
listener does so by responding based on
these relationally framed events.

So they're just suggesting that "relational frames" are discriminative stimuli. As such, the theory seems indistinguishable from Sidman's theory, except the authors go on to confuse me by saying:

The RFT account is different from Sidman’s
equivalence account in four important
ways. First, although Sidman provided one of
the earliest behavioral accounts of stimulus
equivalence, his approach was, and is, primarily
a descriptive one. In fact, he noted,
‘‘My own theorizing has been directed not
so much at an explanation of equivalence
relations but rather, at the formulation of a
descriptive system—a consistent, coherent,
and parsimonious way of defining and
talking about the observed phenomena’’
(Sidman, 1994, p. 536). A precise, coherent
description of empirical phenomena is important,
but it does not satisfy the need for a
functional, behavioral explanation. Sidman’s
account, then, is a description of the behavioral
phenomenon known as stimulus equivalence,
whereas RFT is a behavioral explanation
for how that phenomenon (and other
phenomena) might come about.

where their leading argument against the two approaches being similar is that Sidman claimed that his approach is descriptive whereas RFT is explanatory - the problem being that they are assessing Sidman's theory based on comments and information from four years prior to when he conceptualised his theory... His 1996 paper was part of his preliminary investigation into the issue, and it wasn't until 2000 when he set out the details of his theory and provided the explanatory claims that could be falsified. Sidman's theory cannot be viewed as purely descriptive.

seeker wrote:How would you justify your assessment that “pavlovian conditioning is arguably more important than operant conditioning”? I agree that pavlovian conditioning seems to be a relevant part of the phenomenon, but that´s not enough to say that it´s “more important” than operant conditioning, or that an operant account such as RFT can be dismissed. Also, there´re a lot of empirical and theoretical controversies about the differences and relations between pavlovian and operant conditioning (including some proposals that they´re different experimental conditions with simmilar underlying neural mechanisms), and those controversies should be discused and clarified before arguing such conclusion.

Bah, I thought it was in that Minster and Elliffe paper but the information I was thinking of is from a currently unpublished conference discussion. Basically, they set up an experiment where operant (reinforcement) contingencies and respondent-type stimulus pairings conflicted. So they had a stimulus equivalence situation where the possible choice option was that predicted by previously reinforced trials, and one that had never been reinforced but was associated with the exemplar, and they found near exclusive choice for the respondent relation.

Plus, there's also the experiments that show you can get equivalence without reinforcement:

Saunders, Saunders, Kirby & Spradlin (1988)
Harrison & Green (1990)
Tonneau & González (2004)

So whilst it's not really possible to currently say which process is more important, a reasonable argument could certainly be made that Pavlovian conditioning is more important. The problem this poses for RFT is that its lack of emphasis on the Pavlovian aspect of stimulus equivalence makes it very difficult for it to account for the results we see.

And indeed, I know there is some discussion over the separation of classical and operant conditioning, but it would be up to RFT proponents to make a convincing argument for a collapse of the distinction before we can view it as convincing. As it currently stands, we have documented evidence of equivalence relations being formed with no reinforcement component, so even if they did convince us that we shouldn't view the two as separate processes, their theories would still have some difficulties getting over that hurdle.

seeker wrote:I don´t think so, and we could easily test this claim with a bibliometric assessment of some journals. RFT is not “ignored” in the literature of stimulus equivalence and language. On the contrary, it´s considered one of the most relevant rival accounts (even by its critics, as the ones that you´ve mentioned). The most mentioned accounts of stimulus equivalence and language in the behavior-analytic literature are: RFT, Sidman´s theory, Tonneau´s theory, and Horne & Lowe´s “naming” theory. Also, unlike the other research programs that you´ve mentioned, RFT researchers are studying many of the more relevant aspects of language (they are doing research about rules, instructions, analogies, metaphores, mental concepts, moral concepts, emotional concepts, reasoning, grammar). None of the other research programs are doing studies of so many aspects of language. See also the interesting convergencies between RFT and cognitive research:
http://www.nuigalway.ie/psychology/documents/dale_2002.pdf
http://cognaction.org/rick/pdfs/papers/dale_2004.pdf

I agree that one advantage of RFT is that it does explicitly investigate more areas of language and cognition, but I think this is mostly to do with the way they've proposed their arguments where they've phrased the concepts of stimulus equivalence in a more "cognitive" language that allows for more overlap in the fields. I don't think there is anything special about RFT in that sense though apart from terminology - that is, most of that research you're discussing is made possible through stimulus equivalence, and not RFT.

As for the other theories you mentioned there, "naming theory" never really took off since technically it was disproven the year before they proposed it when Sidman found that naming did not mediate the formation of equivalence. Then later people like Vause, Martin, Yu, Marion and Sakko (2005) demonstrated that even people with minimal verbal skills who did not employ a naming technique during training still learnt equivalence relations.

RFT is not completely discounted, but I still think this is mostly due to the vagueness of the proposal rather than its explanatory power. From the evidence it seems clear that any theory of stimulus equivalence absolutely requires a large emphasis on the role of respondent conditioning and RFT lacks this. Sidman's theory was disproven by the Minster and Elliffe paper, and our best explanation for stimulus equivalence is surely Tonneau's stimulus-relation theory.

I must admit a poverty of knowledge on this. Yet language is something that I need to look at in more detail. I like studying how self organizing mechanistic systems can be used to model more general brain or neural function. Yet language remains pretty much a void. I need to study this stuff.

katja z wrote:I find the technical aspects of this are way over my head, but I'm interested in what a behavioural approach to language can teach linguistics. What I found particulary promising in the RTF is the interest in some difficult aspects of language use - for some reason, I'm especially interested in any work done on the metaphor. Though again, I'm not sure how you can experimentally investigate that, given that it isn't all that clear that it is a discrete category of language use.

I think it's probably important to remember that RFT =/= stimulus equivalence. So stimulus equivalence could provide an explanation for specific language structures like metaphors, and RFT then explains how stimulus equivalence occurs - obviously there will be some overlap where RFT makes specific predictions, but there's probably no particular need to employ RFT before even seeing if metaphor can be explained by stimulus equivalence.

On that note, however, this paper suggests that a metaphor can be understood in stimulus equivalence terms as such:

This interpretation sees the metaphor as involving four elements: (a) establishing two separate equivalence relations, (b) deriving an equivalence relation between these relations, (c) discriminating a formal relation via this equivalence-equivalence relation, and (d) a transformation of functions on the basis of the formal relation discriminated in the third element.

It then goes on to explain how this fits into an RFT framework later in the paper, if you were interested.

katja z wrote:We probably all been taught in school that metaphor and metonymy relie on completely different mechanisms, and it used to be (structuralist) linguistic orthodoxy that they do, but I've always inclined more towards Genette's deconstruction of this opposition, I find his work on this very convincing. Furthermore, when speaking of vocabulary, the line between the literal and metaphoric uses of a word is never easy to draw and it shifts over time, so here again is a source of ambiguity and confusion. I think metaphor-related phenomena are among the crucial challenges for any theory of language, and at the same metaphor may be one of those commonsense concepts that might actually be counterproductive in research if taken at face value, so I'd be interested in seeing how this can be tackled in psychological research and what insights such research can yield for other fields that study aspects of language.

Yeah I imagine the concept of a metaphor is an incredibly difficult one to investigate because, essentially, it's an abstract concept derived from abstract relations... I suppose that a metaphor would probably be best described in stimulus equivalence terms as an inter-concept class relation, where there is a common element between two distinct classes that is linked in some way - presumably, by the specific meaning that someone is trying to get across. So if I said "The moon is a ball of cheese", then the two concept classes would be the moon and cheese, with the linked relation being the holes or craters. In this sense, there's probably no difference between a metaphor and a simile at this level..

katja z wrote:ETA a quick (and probably silly) question on Tonneau, when he speaks of functional equivalence, I suppose this isn't the same as the functional (dynamic) equivalence found in translation studies? (In translation studies, this refers to a focus on content and communicative function of source and target texts, as opposed to formal (word-for-word) equivalence.)

This is how Tonneau explains it:

At the most general level, variables
or procedures are functionally equivalent if they
have the same effect on behavior. Thus, asking
whether the initial links of concurrent
chains are “functionally equivalent” to timeout
periods in multiple schedules (Nevin &
Grace, 2000, p. 81) amounts to asking whether
these factors have the same behavioral effects.
Any adequate theory of behavior should be
able to explain why various procedures or variables
are or aren’t functionally equivalent. Indeed,
searching for functionally equivalent procedures
may be a good way to isolate fundamental
independent variables (e.g., Mazur,
1984).

I don't think the two are directly linked (in that I doubt one was influenced by the awareness of the other approach), but they most likely result from the same thought process. In other words, the subject matter is functionally equivalent (whether it be behaviors or translated words) if they have the same/similar outcome, even if their topography may differ. So a pigeon pecking on an illuminated key and a rat pressing on a lever are engaging in a functionally equivalent behaviors; even though they take different shapes, they're trying to "get the food". I assume this is the same in translation, where even though certain words may not be a like-for-like swap, it's their outcome or overall meaning which is taken into account.

katja z wrote:Damn, I need to get a firmer grip on the terminology of behaviourism. Another foreign lingo to learn ...

To be fair, most behaviorists disagree over what their own terminology means... And then you have people like the ones who created RFT using different words to describe the exact same processes already defined in stimulus equivalence.. It's a wacky world.

katja z wrote:Still, as far as I can see the paper you quote from (I'll look at it a bit later, sorry) uses a definition that fits the classical Aristotelian account of metaphor as a compacted simile. This is fairly restrictive but has the advantage of being well-defined so it can be a useful starting point, but trying to define all instances of metaphor in terms of this structure is problematic. At a guess, there's a confusion stemming from the fact that this is a term from rhetorics and literary criticism adopted into linguistics, and there's no guarantee that all phenomena that have been identified as metaphor operate on the same mechanisms, so what we have here is a bit of a conceptual mess (not so different, I suppose, from what you get when you try to map concepts from folk psychology onto brain processes).

Yeah there will probably be exceptions but I think the key to conducting science in these difficult areas is to look for general trends and try to explain them first, and once you've done that then you try to explain why there are exceptions. So if the definition I presented above is good enough to account for, say, 85-90% of all known metaphors then that would be a good starting point. Then, more than likely, the remaining 10-15% will have some extra rule that governs it which we can hopefully discover by looking at them in terms of our previous explanation.

katja z wrote:Thanks for the clarification on functional equivalence, it seems I wasn't so far off the mark. I don't understand Tonneau's explanation though, whose behaviour is it he talks about in the first sentence, the speaker's or the addressee's?

He's just talking about behavior in choice situations in that explanation, so it's very general and not discussing language there (but the point is that the concept of functional equivalence applies to language in the way he describes). He goes on later in that paper to discuss the more specific form where there is a "transfer of function" like in the standard Pavlovian example - so when we train dogs to salivate to the sound of a bell, then the stimuli "food" becomes functionally equivalent to "bell sound" because they produce the same behavior, "salivation". The point is that this "transfer of function", which results in functional equivalence, is one of the main mechanisms behind stimulus equivalence. The discussion of how this all relates to language is a longer and more complicated one though. (In case all the papers had become confused, this is the paper where he expands on functional equivalence: Equivalence Relations: A Critical Analysis).

Mr.Samsa wrote:Well what is it about RFT that you find so convincing? I've heard of people who find the concept plausible, but I don't know any who seem quite as convinced as you (except perhaps Hayes and Barnes-Holmes, or therapists).

I assess RFT by several criteria in comparison with its rival accounts, and it fares well on those criteria: it has a better empirical evidence, a wider area of application, its research program is more productive, and its comnection with other scientific theories (e.g. neuroscience, cognitive research) are stronger. As you´ve noticed, Sidman´s and Horne & Lowe´s accounts have considerable negative evidence, Tonneau´s program is narrower and less productive, and has remained silent on the most relevant aspects of language. I have my own criticisms to RFT proponents (I disagree with their more philosophical discourses and with some of the terminology that they´ve used), but I still would defend the theory and the research program.

Mr.Samsa wrote:To be honest, this paper just confused me further. This part for example:... So they're just suggesting that "relational frames" are discriminative stimuli.

Where did you read that? No, relational frames are not discriminative stimuli: they are operants controlled by discriminative stimuli.

Mr.Samsa wrote:As such, the theory seems indistinguishable from Sidman's theory, except the authors go on to confuse me by saying...

No, they´re different theories. RFT accepts Sidman´s description of stimulus equivalence, but (1) RFT proposes its own account of how stimulus equivalence is learned (multiple exemplar training) and controlled (discriminative stimuli), and (2) RFT is doing research about many relations that are different from stimulus equivalence (and that Sidman has never studied). Anyway, the differences between RFT and Sidman´s account is not the topic of the article.

Mr.Samsa wrote:Bah, I thought it was in that Minster and Elliffe paper but the information I was thinking of is from a currently unpublished conference discussion. Basically, they set up an experiment where operant (reinforcement) contingencies and respondent-type stimulus pairings conflicted. So they had a stimulus equivalence situation where the possible choice option was that predicted by previously reinforced trials, and one that had never been reinforced but was associated with the exemplar, and they found near exclusive choice for the respondent relation.

I need to read the study to draw any conclusion.

Mr.Samsa wrote:Plus, there's also the experiments that show you can get equivalence without reinforcement...

Yes, I´ve readen them, but they didn´t controlled the pre-experimental history of reinforcement (which is precisely what RFT proposes). The relational frame might have been learned by reinforcement in the pre-experimental history, and it might be evoked by discriminative stimuli during the experiment, without actually having an explicit reinforcement during the experiment. This is the claim of RFT, and this claim hasn´t been refuted yet.

Mr.Samsa wrote:So whilst it's not really possible to currently say which process is more important, a reasonable argument could certainly be made that Pavlovian conditioning is more important.

No, we cannot draw such conclusion when the pre-experimental history is not controlled.

Mr.Samsa wrote:And indeed, I know there is some discussion over the separation of classical and operant conditioning, but it would be up to RFT proponents to make a convincing argument for a collapse of the distinction before we can view it as convincing.

I think that the evidence is the neural research about the role of dopamine-mediated long term potentiation in operant and pavlovian conditioning. By the way, I think that neural research is relevant for solving some behavioral controversies (including the one we were having about molecular-molar research, see my last post about that issue).

Mr.Samsa wrote:As it currently stands, we have documented evidence of equivalence relations being formed with no reinforcement component, so even if they did convince us that we shouldn't view the two as separate processes, their theories would still have some difficulties getting over that hurdle.

There´s no negative evidence until the pre-experimental history is controlled.

Mr.Samsa wrote:I agree that one advantage of RFT is that it does explicitly investigate more areas of language and cognition, but I think this is mostly to do with the way they've proposed their arguments where they've phrased the concepts of stimulus equivalence in a more "cognitive" language that allows for more overlap in the fields. I don't think there is anything special about RFT in that sense though apart from terminology - that is, most of that research you're discussing is made possible through stimulus equivalence, and not RFT.

I don´t think so. RFT has studied many relations that are different to stimulus equivalence. I´ve mentioned some of them: “A is opposite to B” (frame of opposition), “A is different from B” (frame of distinction), “A is bigger than B” (frame of comparison), “A is better than B” (frame of comparison), “A is part of B” (hierarchical frame), “A occurs after B” (temporal frame), “A is condition for B” (conditional frame), “A is mine, B is yours” (deictic frame). It´s clear that none of these relations are “stimulus equivalence” (they don´t comply the criteria of simmetry and transitivity). So you have the burden of proof: how would you explain all those relations just through stimulus equivalence?
RFT has its own proposal to explain all those relations (“each relation is learned through multiple exemplar training, and it becomes a relational operant controlled by discriminative stimuli”). This proposal is not reductible to stimulus equivalence.

Mr.Samsa wrote:RFT is not completely discounted, but I still think this is mostly due to the vagueness of the proposal rather than its explanatory power. From the evidence it seems clear that any theory of stimulus equivalence absolutely requires a large emphasis on the role of respondent conditioning and RFT lacks this. Sidman's theory was disproven by the Minster and Elliffe paper, and our best explanation for stimulus equivalence is surely Tonneau's stimulus-relation theory.

The evidence for pavlovian conditioning is weak until pre-experimental history is controlled (and there´s still the problem of how to discriminate pavlovian vs operant neural mechanisms), so I disagree with your strong emphasis on such issue. Tonneau´s theory has not yet even attempted to formulate hypotheses for most relevant aspects of language, so it would be an overstatement to call it a “theory of language”. So far, I still think that RFT is better than its rival theories.

Mr.Samsa wrote:I think it's probably important to remember that RFT =/= stimulus equivalence. So stimulus equivalence could provide an explanation for specific language structures like metaphors, and RFT then explains how stimulus equivalence occurs - obviously there will be some overlap where RFT makes specific predictions, but there's probably no particular need to employ RFT before even seeing if metaphor can be explained by stimulus equivalence.

I don´t know what do you mean by “RFT =/= stimulus equivalence”, I guess you´re saying “they´re different”. RFT is doing research about many relations, of which stimulus equivalence is only one kind, so it´s true that they´re different theories and research programs. Metaphors rely on many kinds of relations that are different from stimulus equivalence, and nobody has reduced all those relations to stimulus equivalence, so if you´re proposing that “metaphor can be explained by stimulus equivalence”, you´ll have that burden of proof: show us how the other relations are based on stimulus equivalence.

Mr.Samsa wrote:On that note, however, this paper suggests that a metaphor can be understood in stimulus equivalence terms as such:

This interpretation sees the metaphor as involving four elements: (a) establishing two separate equivalence relations, (b) deriving an equivalence relation between these relations, (c) discriminating a formal relation via this equivalence-equivalence relation, and (d) a transformation of functions on the basis of the formal relation discriminated in the third element.

It then goes on to explain how this fits into an RFT framework later in the paper, if you were interested.

That´s only one possibility of analogy: equivalence between two equivalences. That´s possibly the most usual kind of analogy, but you can have analogies between other kinds of relations, e.g. “A is opposite to B, as C is opposite to D”, where the relation “opposite to” doesn´t comply the transitivity criterion of stimulus equivalence. RFT is doing research about those other kinds of relations.

Mr.Samsa wrote:To be fair, most behaviorists disagree over what their own terminology means... And then you have people like the ones who created RFT using different words to describe the exact same processes already defined in stimulus equivalence.. It's a wacky world.

It´s false that the ones who created RFT are “using different words to describe the exact same processes”. They´ve proposed “mutual and combinatorial entailment” as wider concepts, that include the equivalence criteria (simmetry and transitivity) as one of their cases. So they´re using different words to describe different processes.

seeker wrote:I assess RFT by several criteria in comparison with its rival accounts, and it fares well on those criteria: it has a better empirical evidence, a wider area of application, its research program is more productive, and its comnection with other scientific theories (e.g. neuroscience, cognitive research) are stronger. As you´ve noticed, Sidman´s and Horne & Lowe´s accounts have considerable negative evidence, Tonneau´s program is narrower and less productive, and has remained silent on the most relevant aspects of language. I have my own criticisms to RFT proponents (I disagree with their more philosophical discourses and with some of the terminology that they´ve used), but I still would defend the theory and the research program.

Well what empirical evidence is there that specifically supports RFT above and beyond stimulus equivalence?

seeker wrote:

Mr.Samsa wrote:To be honest, this paper just confused me further. This part for example:... So they're just suggesting that "relational frames" are discriminative stimuli.

Where did you read that? No, relational frames are not discriminative stimuli: they are operants controlled by discriminative stimuli.

It was this section: "Accordingly, the definition of verbal stimuli is ‘‘stimuli that have their effects because they participate in relational frames’’" - so the verbal stimuli (which are discriminative stimuli themselves) are a product of the relational frame, which would make the relational frame a higher order discriminative stimuli. It's probably what the applied behavior analysts would call a "setting event".

seeker wrote:

Mr.Samsa wrote:As such, the theory seems indistinguishable from Sidman's theory, except the authors go on to confuse me by saying...

No, they´re different theories. Anyway, he differences between RFT and Sidman´s account is not the topic of the article.

Whether it's the topic of the article isn't important, they propose the same mechanism behind stimulus equivalence, which is that it's a product of operant processes. The specific relations and frames aren't really important to consider since Sidman's theory would make the same predictions.

seeker wrote:

Mr.Samsa wrote:Plus, there's also the experiments that show you can get equivalence without reinforcement...

Yes, I´ve readen them, but they didn´t controlled the pre-experimental history of reinforcement (which is precisely what RFT proposes). The relational frame might have been learned by reinforcement in the pre-experimental history, and it might be evoked by discriminative stimuli during the experiment, without actually having an explicit reinforcement during the experiment. This is the claim of RFT, and this claim hasn´t been refuted yet.

Mr.Samsa wrote:So whilst it's not really possible to currently say which process is more important, a reasonable argument could certainly be made that Pavlovian conditioning is more important.

No, we cannot draw such conclusion when the pre-experimental history is not controlled.

Of course it hasn't been controlled, it's impossible to control for it... If that's what you need in order to disprove RFT then you're essentially claiming that it is unfalsifiable (since animals cannot do stimulus equivalence, and arguably never will be able to, so we can't control the reinforcement histories of human subjects).

seeker wrote:

Mr.Samsa wrote:As it currently stands, we have documented evidence of equivalence relations being formed with no reinforcement component, so even if they did convince us that we shouldn't view the two as separate processes, their theories would still have some difficulties getting over that hurdle.

There´s no negative evidence until the pre-experimental history is controlled.

It doesn't have to be negative evidence, it just needs to be evidence that better fits one theory over another. One theory predicts this result, and the other can potentially deal with it by referring to hypothetical unknown reinforcement histories. It's not negative evidence exactly, but it's a black mark against RFT.

seeker wrote:

Mr.Samsa wrote:I agree that one advantage of RFT is that it does explicitly investigate more areas of language and cognition, but I think this is mostly to do with the way they've proposed their arguments where they've phrased the concepts of stimulus equivalence in a more "cognitive" language that allows for more overlap in the fields. I don't think there is anything special about RFT in that sense though apart from terminology - that is, most of that research you're discussing is made possible through stimulus equivalence, and not RFT.

I don´t think so. RFT has studied many relations that are different to stimulus equivalence. I´ve mentioned some of them: “A is opposite to B” (frame of opposition), “A is different from B” (frame of distinction), “A is bigger than B” (frame of comparison), “A is better than B” (frame of comparison), “A is part of B” (hierarchical frame), “A occurs after B” (temporal frame), “A is condition for B” (conditional frame), “A is mine, B is yours” (deictic frame). It´s clear that none of these relations are “stimulus equivalence” (they don´t comply the criteria of simmetry and transitivity). So you have the burden of proof: how would you explain all those relations just through stimulus equivalence?
RFT has its own proposal to explain all those relations (“each relation is learned through multiple exemplar training, and it becomes a relational operant controlled by discriminative stimuli”). This proposal is not reductible to stimulus equivalence.

Those are just specific relations which are a product of the mechanism that RFT proposes, they aren't the subject matter that is under question though. It's the mechanism that the RFT proponents are suggesting controls stimulus equivalence that is important. I can't find how RFT differs from Sidman's theory in this regard, and as such his theory would make the same predictions about the relations you've described above (assuming they have evidence backing them up).

Can you link me to some studies that demonstrate the prediction and control of these relations? (Or refer me to which papers you've linked to might be relevant). I'm trying to figure out how exactly they study them.

seeker wrote:I don´t know what do you mean by “RFT =/= stimulus equivalence”, I guess you´re saying “they´re different”. RFT is doing research about many relations, of which stimulus equivalence is only one kind, so it´s true that they´re different theories and research programs.

Yes, I was pointing out that they are different.

seeker wrote:Metaphors rely on many kinds of relations that are different from stimulus equivalence, and nobody has reduced all those relations to stimulus equivalence, so if you´re proposing that “metaphor can be explained by stimulus equivalence”, you´ll have that burden of proof: show us how the other relations are based on stimulus equivalence.

You've kind of taken my quote out of context there, I said: "...obviously there will be some overlap where RFT makes specific predictions, but there's probably no particular need to employ RFT before even seeing if metaphor can be explained by stimulus equivalence". In other words, since RFT relies on stimulus equivalence principles and then adds its own assumptions and assertions, it makes sense to analyse metaphors in terms of stimulus equivalence first to see where the evidence takes us. The evidence might support RFT, I don't know, but the point is that we should be following the evidence to draw conclusions about the appropriate method of analysis, rather than using one specialised approach to analysing metaphor at the cost of other approaches.

seeker wrote:That´s only one possibility of analogy: equivalence between two equivalences. But you can have equivalence between other relations, e.g. “A is opposite to B as C is opposite to D”, where the relation “opposite to” doesn´t comply the transitivity criterion of stimulus equivalence.

I imagine that Tonneau would probably argue that it's a generalised operant, where basically the rule is transferred. (Essentially, the explanation would be the same as RFT).

History of reinforcement can be a garbage dump if you are not careful, but an advantage of it is that you can manipulate it. The RFT researchers have done that in many studies, and not just with equivalence (although there too, eg infants -- see Luciano, M. C., Gómez, I., & Rodríguez, M. (2007). The role of multiple-exemplar training and naming in establishing derived equivalence in an infant. Journal of Experimental Analysis of Behavior, 87, 349-365.)

For example here is one with comparative relations:

Berens, N. M., & Hayes, S. C. (2007). Arbitrarily applicable comparative relations: Experimental Evidence for relational operants. Journal of Applied Behavior Analysis, 40, 45-71. In a combined multiple baseline (across responses and participants) and multiple probe design (with trained and untrained stimuli), this study found that reinforced multiple exemplar training facilitated the development of arbitrary comparative relations, and that these skills generalized not just across stimuli but also across trial types.

Here is another, with deictic relations (I/you, here/there, now/then) and theory of mind skills:
Weil, T. M., Hayes, S. C., & Capurro, P. (in press). Establishing a deictic relational repertoire in young children. The Psychological Record.

There are a half a dozen others

As for the idea that multiple stimulus relations can be accounted for by equivalence and thus by Sidman's account -- that idea simply will not work. No one has shown experimentally that it will, and I know of no major figure even making that claim. if I can appeal to authority, here is a quote by Murray Sidman in his 2008 article in Cognitive Studies:

[my approach is] “a limited theory
in that it does not cover other kinds of relations than equivalence, as for example,
relational frame theory attempts to do” (2008, p. 331, emphasis in the original).

Murray has repeatedly and explicitly denied that his approach is an attempt to construct a comprehensive
approach to language and cognition. In that context it's a bit jarring to see the posts that seem to treat equivalence as an adequate approach to many areas in language and cognition, justifying that claim by an appeal to Sidman.

That idea is a dead horse. Empirically, it certainly seems to be. Equivalence is 40 years old and has roots that go back several decade more. If that idea is so sound, where are the data?

RFT is trying to build a comprehensive contextual behavioral approach to language and cognition.
Although I think the RFT account of equivalence is
doing fine, no one has done studies on metaphor, sense of self, implicit cognition, complex forms of rule-governance,
(etc etc) using just equivalence -- and RFT is deep into such areas and many more (as some of the some on this list have pointed out). Thus it is apples and oranges to look at RFT as if it is a theory of equivalence. It is far more than that and if you want to criticize RFT a more productive angle is to get deeply into a given domain (pick one: sense of self and perspective taking; metaphor and analogy; implicit cognition; etc) and see how well the theory is doing.

There are lots of things not yet established -- that is true -- but there are no empirical sours note that I know of. If you want to deal with criticisms there are lots of things to look at and argue about. That is fine, fun, and potentially useful .. but RFT folks have hardly been silent in addressing criticisms. If you want to take these criticisms seriously can't just read the critics. You need then to go back and carefully read the original material and the studies. Very often fair readers will see in just a few hours that critics are selectively quoting; misquoting; inserting philosophical biases; ignoring data; and so on. To this moment not a single criticism has legs that can be turned into studies that I know of. If you disagree, please give me the criticism you think is telling and how it would be tested.

But there is another kind of response as well. I like the post that said basically "we need to look at evidence of progressivity and productivity." Right on. We do. RFT is sitting on about 150 studies (there is a list on the website ... address below) and more come out every month in a very wide range of areas. If you read the actual studies many of them are dealing with predictions no one else was making at the time (start with Steele and Hayes, 1991 -- the first real 100% RFT. None of the many behavior analysts I asked at the time predicted those outcomes. Not one.). And then look over time in the way that the research program has lead to new things (e.g., deictic framing and sense of self; theories of implicit cognition). And don't forget application ... not just in therapy settings but in assessment, language training, and other areas. That is important in my book.

Then go look to see whether the raps by Burgos, Tonneau etc have lead to a sequence of studies that gradually open up new areas.

I don't expect me saying such things to be convincing in and of themselves -- but the data should be (and if they are not
then honest critics should say what data are missing). That takes work ... but that is just the way it is in science.
Statements like "Well what empirical evidence is there that specifically supports RFT above and beyond stimulus equivalence?" are kind of stunning. This is a vast literature and very little of it has to do with equivalence. "Prove to me its not equivalence" flips science on its head. It's your job to provide the account if you think you can, not my job to disprove your theory you've not put into written or empirical form.

You want single study to try your equivalence based account on? OK. Look at Dougher, M. J., Hamilton, D., Fink, B., & Harrington, J. (2007). Transformation of the discriminative and eliciting functions of generalized relational stimuli. Journal of the Experimental Analysis of Behavior, 88(2), 179-197. This study examines the transformation of respondent and operant stimulus functions via more-than/less-than direct and derived relations.
If you think you can do an equivalence based account -- have at it.

Here is another

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. Erratum, 66, 348.
The first study to show three patterns of derived relational responding in accordance with sameness, more-than, and less-than. Alternative explanations for the transformation test outcomes (e.g., based on equivalence) are considered and found wanting.


If you want to access RFT studies easily, the cheap way is to join ACBS (costs a buck) at www.contextualpsychology.org and download the articles. Also you can then join a list serve with several hundred RFT interested folks on it ... any criticism etc posted there will generate intelligent and thoughtful responses.

Hope I have not intruded.

Steve Hayes

Very excellent 1st post Steve and to the intrusion

I've played with self organizing mechanistic systems in attempts to improve neural net technologies beyond the limits they now have. Especially with respect to predefined input/output directionality, not the path which the standard approach doesn't specify. Though I think I have some good clues on the basics, more complex AI intelligences will require a far more open ended approach to the self organization of intelligences. My limited knowledge and sense of RFT is that it holds some promising ideals that might be formulated in algorithms. Self organized systems remains a nascent math and science, but I like this field. Biological and psychological sciences are a nice source of raw data.

Your rebuttal to the framing of RFT as an equivalence theory appears reasonably convincing to me. What studies and/or experiments do you think would be the most useful for someone wanting to consider it in terms of algorithm development?