I have read and reread (and reread again!) chapter 2 over the past month, and find it much denser even than chapter 1. The concept I get hung up on over and over is the notion of arbitrary applicability (AA). While I know that AA is defined as a relationship that depends on social whim or consensus instead of formal characteristics of the stimulus, I think I’m missing the implications.

If we look at the equation for Cfunc as expressed on page 33, there’s a sentence that follows that says: “[w]e can say it this way: given arbitrarily applicable stimulus relations between A, B, and C, and given a context that actualizes the transformation of a given function of A, the functions of B and C will be modified in terms of the underlying relations between A, B, and C.”

I’d like to substitute examples for “A, B, and C”—could someone tell me if I’m on the right track with this?
An example of an actualizing context would be people talking to each other, having a conversation.
In that context, A might be a banana (the fruit itself, not the word).
One person could say, “Have you ever noticed how her nose [B, in my example?] looks like a banana [the oral word for the fruit, C in my example?]?”
Now, this comparison relies on both people’s previous experience with bananas. If the person responded “What’s a banana?” the first person would have to whip out a banana, or at least a picture of a banana, for purposes of comparison.
If this is right so far, one question would be, does one of the 3 items in the relation have to be something with material existence, not “just” a word? I think the answer to this is “no” because what if my characters were talking about something abstract?

So . . .
One person might say, “Love [A] is blind [B].”
Would C in this example be the quality of not being sighted, of blindness?
I’m getting tangled up here. Help!

Response Submitted

Funcky Cstuff. (Submitted by Jacqueline A-Tjak on August 1, 2006 - 2:55am.)

Hi Leslie,

I hope I do this right and this message will appear as an reply to your post.
The example you give of nose and banana seems right to me. The function: 'bended' tranfers/transforms from the actual banana to the nose (word).
Now the second example. Here is a 'simple' way to look at this: I do not know whether you need the third part (C). Transfer of function can happen too when it is about two relata, A en B, Love and blindness. There would be a transfer of the function: not seeing properly from the word blind to the word love.

Now lets make it more complicated:
The word blind refers to actually being blind or seeing someone that is blind (modeling). You could understand the sentence 'love is blind' without ever having seen that, just by hearing people tell about it.
With 'love' its even more complicated, because that is an even more complex experience.

Your question was: does one of the 3 items in the relation have to be something with material existence, not “just” a word.
The answer, as far as I understand it, is no, but....Somewhere in the learning history has to be direct experience. If you take the example of numbers: They are abstract, but we learn them by counting actual things. So your example would not be about three relata, but maybe thirty relata that add up to 'love is blind'. And one or more of these relata would be derived from experience.
And I think it would be far to complicated to try and figure out what relata you need to get at 'love is blind'.
So for the sake of understanding the matter maybe it is easier and enough to stay with the two relata.

Well, I hope this makes sense and it is a good test to see if I get what RFTstuff is about.

Jacqueline

On page 30, first full paragraph, there is a description of a natural language event--someone names a ball to a child. It's being used to illustrate mutual entailment. But it seems that the last sentence "the r response in other words, will involve responding to the sound "ball" in terms of the previously experienced functions of actual balls." So that seems to be a description of the transfer of stimulus functions.

Is transfer of stimulus function a more precise term than transformation of stimulus function. Is transfer of stimulus function a subset of transformation of stimulus function that applies with frames of coordination or is it a different thing altogether?

Joanne Steinwachs

Leslie Telfer

Response to initial post-

Just go on reading... (Submitted by Philippe Vuille on August 5, 2006 - 9:28am.)

The response to your question is to be found on pages 31-32 :

«Equivalence research has repeatedly revealed that stimulus functions commonly transfer through the members of equivalence classes.»

Consider the classical example given by Catania (Learning, 4th ed. p. 154) :

«A child has learned to obey a parent's words, go and stop, when crossing with the parent at a traffic intersection. In a separate setting, the child is taught that go and green traffic lights are equivalent and that stop and red traffic lights are equivalent (in other words, go and green become members of one equivalence class and stop and red become members of another). If the discriminative functions of the words go and stop transfer to the respective traffic lights, the child will obey the traffic lights without any additional instruction.»

RFT is not only about equivalence or "coordination frames". In the case of frames such as opposition, the stimulus functions will not be merely transferred but transformed, as is shown in the example on page 32 :

«(...) Suppose a person is trained to selext stimulus B as the "opposite" of stimulus A. Now suppose that A is given a conditioned punishing function, such as by pairing it with a loss of points. It might be predicted that B would then have reinforcing functions (without having that function directly trained), by virtue of its "opposite" relation to the punishing A stimulus (...) It hardly seems right to say that the reinforcing effects "transferred" in such a case, because they were acquired indirectly through the relation of opposition between B and a punisher. It seems more proper to use the term transformation than transfer, and it is for this reason that RFT has adopted transformation of stimulus functions as the general term for this effect. We will still use the term transfer of stimulus functions, but will generally reserve it for situations in which the underlying relation leads to derived functions that are similar to those that were trained or that pre-existed.»

Philippe

Hi all,

Here are some questions I have after reading chapter 2.

1. In paragraph 2.1.1 (page 22/24) is explained what overarching purely functional operants are. Could one say that 'avoidance' is an overarching purely functional operant?

2. Does anyone know what 'self-discrimination functions' are? They are mentioned on page 32 in a paragraph about tranformation of stimulusfunctions (2.2.3)as an example of stimulusfunctions that have been shown to transfer.
As a whole, I still find the concept of stimulusfunctions very difficult to grasp. What I find especially difficult is to find good examples and to specify stimulusfunctions that are involved when looking at a real life example. Anyone who can help out here with examples?

3. In paragraph 2.4 families of relational frames are summed up (page 35-39). If I get it right the phrase: 'snakes are dangerous' means NOT that the snakes are in a relation of coordination, which should be understood as equivalence, but in a relation of hierarchie, like the phrase: John is an man. Snakes are a part of dangerous stuff. Is this right?

4. Am I right that stimulusfunctions can be relata?

Jacqueline

A discussion of this chapter is included below:

re: Some questions. (Submitted by JT Blackledge on September 4, 2006 - 11:02pm.)

Hi Jacqueline--

Good questions--I've inserted some answers below.

1. In paragraph 2.1.1 (page 22/24) is explained what overarching purely functional operants are. Could one say that 'avoidance' is an overarching purely functional operant?

I think it would be slightly more technically correct to term avoidance behaviors as members of the same functional class--as avoidance behaviors have over the relatively long history of behavior analysis. Reserving the word 'overarching' for language (for example) allows the term to illustrate how operants can be grouped together even if these operants involve different functions.
(this answer might make more sense after reading the answer to 2, below).

2. Does anyone know what 'self-discrimination functions' are? They are mentioned on page 32 in a paragraph about tranformation of stimulusfunctions (2.2.3)as an example of stimulusfunctions that have been shown to transfer.
As a whole, I still find the concept of stimulusfunctions very difficult to grasp. What I find especially difficult is to find good examples and to specify stimulusfunctions that are involved when looking at a real life example. Anyone who can help out here with examples?

A stimulus function essentially refers to three things: the stimulus itself, the response made with respect to it, and the [punishing or reinforcing] consequence received for making that response. For example, you could say that a stimulus (e.g., a snake) serves an avoidance function if the presentation of a snake leads a subject to run away, and this 'running away' response is negatively reinforced by the 'removal' of the snake. Or, you could say that a child screaming serves an attention function if the screaming--made in response to an adult previously ignoring him--results in the adult positively reinforcing the child by attending to him. Finally, a 'self-discrimination function' might be evident if the verbal stimulus "Where are you, Jacqueline?" resulted in you responding "I'm right here", which in turn resulted in me positively reinforcing you. Essentially then, a "stimulus function" refers to a stimulus, a response made to that stimulus, and a consequence for that response. The 'whole behavioral unit', so to speak.

I've noticed the term "stimulus function" used more loosely & informally. For example, one might say that film of a combat scene serves an "emotive function", which would imply that the stimulus (the film) makes the viewer feel a certain way. Or, the term 'experiential avoidance' might be used to describe an attempt to avoid an unpleasant feeling that doesn't 'succeed' (i.e., doesn't yield negative reinforcment by immediate termination of the unpleasant feeling). Not entirely technically correct usage, as instances like this don't specify the consequence to the response-- though it does describe, in varying detail, the topographical nature of the response.

3. In paragraph 2.4 families of relational frames are summed up (page 35-39). If I get it right the phrase: 'snakes are dangerous' means NOT that the snakes are in a relation of coordination, which should be understood as equivalence, but in a relation of hierarchie, like the phrase: John is an man. Snakes are a part of dangerous stuff. Is this right?

Correct. The critical thing to think about when deciding if two stimuli are in a hierarchical relation vs. a relation of coordination is this: Does 'reversing' the relation make sense? If it's a coordinative relation, it will make logical sense without changing the relational term (e.g., changing "noir is black" to "black is noir" makes sense, as does changing "pictures are snapshots" to "snapshots are pictures"). If it's a hierarchical relation, it won't make sense: changing either "snakes are dangerous" to "dangerous are snakes" or "dogs are mammals" to "mammals are dogs" doesn't work, since in each case, one of the relata are members of a larger class and do not exclusively or exhaustively define the class.

4. Am I right that stimulusfunctions can be relata?

I think they would be more technically referred to as stimuli, but this relates to your question #2. Within behavior analysis, a "stimulus function" refers to a stimulus, the response made with respect to that stimulus, and the consequence that follows. If you think of relata as stimuli, and grant that (within behavior analysis) stimuli are typically only talked about if they have a function (why include a stimulus in a functional analysis unless it affects subsequent behavior?), then it follows that, for practical purposes, every stimulus has a function--and that any relata included in a meaningful functional analysis are thus not only stimuli, but also partipate in stimulus functions.

Best,
JT

JT Blackledge
Lecturer
University of Wollongong
New South Wales, Australia

(Inevitably) some more questions. (Submitted by Jacqueline A-Tjak on September 5, 2006 - 7:09am.)

Hi JT

Thank you do much. Your answers made things more clear to me. And led to more questions.
In chapter one it is metioned that 'stimulusfunctions could be substitutive, no longer requiring the presence of a stimulusobject' (page 8. This is why I thought stimulusfunctions can be relata or stimuli on their own right. After reading your postand thinking it over I now come to the conclusion that there must be some kind of stimulus, non-verbal or verbal that 'contains' the stimulusfuntion and stimulusfunctions do not exist in a void. Do I understand this properly this way?

And do you mean to say that self-discrimination functions mean that a person discriminates himself in a way?

Now, you put stimulusfunctions in a frame of operant conditioning, if I may put it this way. But on page 31 the RFT book says that elicited conditioned emotional responses have been proven to be stimulusfunctions that transfer. Do you think/know, does this mean that these responses must be understood in a operant context, for instance that they are functioning as a discriminative stimulus? Or do stimulusfunctions also operate in a classical conditioning way?
You write:
I've noticed the term "stimulus function" used more loosely & informally. For example, one might say that film of a combat scene serves an "emotive function", which would imply that the stimulus (the film) makes the viewer feel a certain way. Or, the term 'experiential avoidance' might be used to describe an attempt to avoid an unpleasant feeling that doesn't 'succeed' (i.e., doesn't yield negative reinforcment by immediate termination of the unpleasant feeling). Not entirely technically correct usage, as instances like this don't specify the consequence to the response-- though it does describe, in varying detail, the topographical nature of the response.

Now what you write makes me think we should not view stimulusfunctions as something that happens in a classical conditioning situation, since there are no consequences in purely classical conditioning. I am not sure if you can, in ordinary life (as opposed to the laboratory) divide operant and classical conditioning, but I would like to make sure that I understand it correctly, if that is possible.

Lastly, you answered me that avoidance should be seen as a functional class, not an overarching operant. I am afraid that I do not understand why it is the first AND not the latter.

Thanks for having taken the trouble to clear things up for me and I hope you or somebody else can help me out with these new questions.

Best wishes

Jacqueline

RE: additional questions. (Submitted by JT Blackledge on September 5, 2006 - 9:14pm.)

Hi Jacqueline--
I've inserted answers below in ALL CAPS.

In chapter one it is metioned that 'stimulusfunctions could be substitutive, no longer requiring the presence of a stimulusobject' (page 8. This is why I thought stimulusfunctions can be relata or stimuli on their own right. After reading your postand thinking it over I now come to the conclusion that there must be some kind of stimulus, non-verbal or verbal that 'contains' the stimulusfuntion and stimulusfunctions do not exist in a void. Do I understand this properly this way?

YES--I CAN DEFINITELY SEE HOW THAT SENTENCE IS MISLEADING. THE 'UNWRITTEN SUBTEXT' THERE IS THAT THE STIMULUS IS NOW NOT AN ACTUAL LIGHT, BUT A VERBAL STIMULUS LIKE "IMAGINE A LAMP SITTING IN FRONT OF YOU". THE SENTENCE MIGHT MORE ACCURATELY READ "STIMULUS FUNCTIONS COULD BE SUBSTITUTIVE, NO LONGER REQUIRING THE PRESENCE OF THE ORIGINAL STIMULUS OBJECT".

And do you mean to say that self-discrimination functions mean that a person discriminates himself in a way?

YES

Now, you put stimulusfunctions in a frame of operant conditioning, if I may put it this way. But on page 31 the RFT book says that elicited conditioned emotional responses have been proven to be stimulusfunctions that transfer. Do you think/know, does this mean that these responses must be understood in a operant context, for instance that they are functioning as a discriminative stimulus? Or do stimulusfunctions also operate in a classical conditioning way?

You write:
I've noticed the term "stimulus function" used more loosely & informally. For example, one might say that film of a combat scene serves an "emotive function", which would imply that the stimulus (the film) makes the viewer feel a certain way. Or, the term 'experiential avoidance' might be used to describe an attempt to avoid an unpleasant feeling that doesn't 'succeed' (i.e., doesn't yield negative reinforcment by immediate termination of the unpleasant feeling). Not entirely technically correct usage, as instances like this don't specify the consequence to the response-- though it does describe, in varying detail, the topographical nature of the response.

Now what you write makes me think we should not view stimulusfunctions as something that happens in a classical conditioning situation, since there are no consequences in purely classical conditioning. I am not sure if you can, in ordinary life (as opposed to the laboratory) divide operant and classical conditioning, but I would like to make sure that I understand it correctly, if that is possible.

VERY NICE CATCH OF A VERY FINE-GRAINED POINT. THINK OF IT THIS WAY: A GIVEN STIMULUS CAN TAKE ON DIFFERENT FUNCTIONS DUE TO VARIOUS PROCESSES--CLASSICAL CONDITIONING, OPERANT CONDITIONING, STIMULUS GENERALIZATION, DERIVED RELATIONAL RESPONDING. ONCE A STIMULUS HAS A FUNCTION, REGARDLESS OF THE PROCESS(ES) INVOLVED IN TAKING ON THAT FUNCTION, THAT FUNCTION STILL TECHNICALLY REFERS TO THAT STIMULUS, THE RESPONSE THAT IS MADE TO THAT STIMULUS, AND THE CONSEQUENCE THAT FOLLOWS.

YOU'RE DEFINITELY RIGHT ABOUT THE ARBITRARY DISTINCTION BETWEEN CLASSICAL AND OPERANT CONDITIONING. INDEED, IT'S BEEN ARGUED THAT AN OPERANT CAN BE DESCRIBED IN PURELY RESPONDENT TERMS (I.E., THE DISCRIMINATIVE STIMULUS IS CLASSICALLY CONDITIONED TO THE RESPONSE, WHICH IN TURN IS CLASSICALLY CONDITIONED TO THE CONSEQUENCE). IT'S JUST THE PRACTICAL UTILITY IN DISTINGUISHING BETWEEN TWO PROCESSES THAT KEEPS THEM SEPARATE, METHINKS.

Lastly, you answered me that avoidance should be seen as a functional class, not an overarching operant. I am afraid that I do not understand why it is the first AND not the latter.

IN APPLIED BEHAVIOR ANALYSIS, IT'S COMMMON TO TALK ABOUT BEHAVIOR HAVING ONE OF FOUR FUNCTIONS: AVOIDANCE (OF AVERSIVE STIMULATION), ACQUISTION (OF TANGIBLE REINFORCERS), ATTENTION (I.E., BEHAVIOR THAT FUNCTIONS TO RECEIVE POSITIVE REINFORCEMENT IN THE FORM OF ATTENTION), AND SELF-STIMULATION. ANY BEHAVIOR, REGARDLESS OF HOW DIFFERENT THOSE BEHAVIORS LOOK, THAT SERVES ONE OF THOSE FUNCTIONS (E.,G., AN AVOIDANCE FUNCTION) FALLS INTO THAT RESPECTIVE FUNCTIONAL CATEGORY. IN OTHER WORDS, BEHAVIORS ARE GROUPED INTO FUNCTIONAL CLASSES SOLELY ACCORDING TO FUNCTION--A FUNCTIONAL CLASS, BY DEFINITION, INCLUDES A VARIETY OF STIMULUS/RESPONSES THAT ALL SERVE PRECISELY THE SAME FUNCTION.

OVERARCHING OPERANTS--LIKE MUTUAL/COMBINATIORIAL ENTAILMENT--INVOLVE PROCESSES THAT OPERATE REGARDLESS OF WHAT SUBSEQUENT FUNCTIONS RESULT. IN ANY GIVEN CASE OF COMBINATORIAL ENTAILMENT, FOR EXAMPLE, THE RESULTING FUNCTION MIGHT BE AVOIDANCE, ACQUISITION, ATTENTION, OR SELF-STIMULATION. THINK OF AN OVERARCHING OPERANT AS A PROCESS THAT ALLOWS VARIOUS FUNCTIONS TO 'ATTACH' THEMSELVES (SO TO SPEAK) TO STIMULI IN A CLEARLY SPECIFIED WAY--AND FUNCTIONAL CLASSES AS GROUPINGS OF STIMULUS/RESPONSES THAT SHARE THE SAME FUNCTION (REGARDLESS OF THE PROCESSES THROUGH WHICH THESE FUNCTIONS AROSE).

Thanks for having taken the trouble to clear things up for me and I hope you or somebody else can help me out with these new questions.

NO PROBLEM--IT STRETCHES MY BRAIN :)

Best wishes

Jacqueline

JT Blackledge
Lecturer
University of Wollongong
New South Wales, Australia

One addition. (Submitted by JT Blackledge on September 6, 2006 - 1:11am.)

Actually, you know what, the more I think about it, the more I recall that it would be correct to talk about a respondent function as including only the classically conditioned stimulus and the response made with respect to it (theoretically speaking, no consequence). That is, indeed, the 'whole behavioral unit', when speaking about a respondent, so why not?

JT Blackledge
Lecturer
University of Wollongong
New South Wales, Australia

More brainstretching exercises. (Submitted by Jacqueline A-Tjak on September 6, 2006 - 3:22am.)

Hi JT,

Thanks for clearing things for me and in the proces of doing so make me feel that I am asking questions that are worth answering. Asking these kind of questions always make me feel awkward, since my mind produces a lot of chatter in the proces.

Now if I understand everything you have written, it follows that the phrase: "one might say that film of a combat scene serves an "emotive function", which would imply that the stimulus (the film) makes the viewer feel a certain way (I quoted you here) is actually in accordance with the way stimulusfunctions work. What the speakers means/could mean is that the movie by classical conditioning or derivation has acquired a conditional emotional responsfunction. Is this correct?

Now here comes a difficult question for me to ask:
You wrote a very fine article about RFT: Blacklegde J.T. (2003) An Introduction to Relational Frame Theory: Basics and Applications. (The Behavior Analyst Today, 3, 4, 421-433). I recommend it to everyone who wants to know more about RFT. Soon I will be giving a course in ACT and tell something about RFT and your article is among the literature the coursemembers have to read.
AND..... You say in this article that Snakes and Danger are in a relation of coordinance. So, would this mean you made a mistake there?

Kind regards,

Jacqueline

re: stretching. (Submitted by JT Blackledge on September 6, 2006 - 6:22pm.)

Yes indeed--danger would stand in a hierarchical relation to snakes (not a coordinative one), wouldn't it?

JT Blackledge
Lecturer
University of Wollongong
New South Wales, Australia

Stimulus Functions. (Submitted by Jacqueline A-Tjak on September 17, 2006 - 9:21am.)

Hi all,

Still strugling with the concept of stimulusfunctions (the RFT book does not give a proper definition, does it?) I would like to know if one could say the following to be technically correct:
"Stimulusfunctions specify the nature of influence (not the intensity of it) of a certain stimulus to a certain respons". For instance, a stimulus could have a discriminative function, which would mean that it signals to a person that certain behavior has a acceptable probability to be reinforced. Since we are talking about probabilities you cannot say the stimulusfunction specifies the influence, only the nature of the influence (we are never sure beforehand this influence actually will show itself).

Jacqueline