[Author’s Note, June 2020: This post is now out of date.  It was written before the emergence of crucial new empirical evidence from Jaswal et al. that changed the RPM debate.  I also feel that there are sections of this post that are not sufficiently nuanced given the complexity of this issue and importance of the right to communication.  The post thus does not reflect my current perspective, which can be seen here: http://www.autisticscholar.com/rpm-and-fc/.]

In recent years, a large segment of the autistic community appears to have embraced the highly controversial Rapid Prompting Method (RPM).  RPM is a technique which ostensibly aims to help non-speaking people with disabilities communicate.  Unfortunately, there are reasons to doubt the effectiveness of this approach, which I will describe more fully below.

These reasons have not stopped some autistic advocates from adopting the position that RPM’s claim to support autistic communication must be taken at face value, lest autistic people be denied access to communication.  Others seem to have adopted the somewhat subtler position that the efficacy of communication supports should be judged on an individual basis, and that autistic people should have the right to the communication approach that works best for them – even if that method is RPM.  I am hoping that the recent Autistic Self Advocacy Network (ASAN) open letter on the subject can be interpreted to reflect the latter, subtler view.

However, both of these positions are opposed to the views of many researchers and professionals, whose views are perhaps expressed in a recent draft position statement from the American Speech-Language-Hearing Association (ASHA).  The ASHA statement not only recommends against use of RPM, but adopts the position that communication through RPM should not be regarded as originating from the person with a disability.

So, What is RPM?

RPM has sometimes been compared to facilitated communication (FC), which is now widely regarded as bogus.  In FC, a facilitator holds the arm or hand of a person with a disability while they type on some kind of keyboard or interface.  In theory, the facilitator simply steadies the arm of the individual and helps them complete motor movements.  In practice, we’re now pretty sure that FC doesn’t really work and that it is the facilitator who is often responsible, through moving the hand of the person with a disability, for any communication that occurs.

RPM is a bit different.  In RPM, the aide doesn’t physically touch or guide the arm of a person with a disability.  Instead, the aide holds a letter board in front of the person with a disability while providing continuous rapid prompts to encourage them to point to appropriate letters.  Supporters of RPM discourage testing of the technique, pointing to a previous study as evidence of effectiveness (although this previous study was not intended to, and does not, test whether RPM reflects valid communication).

Unfortunately, it has sometimes been suggested that the only problem with RPM is that it is untested or inadequately tested.[1]  This claim is not accurate, as a basic examination of the approach and its claims illustrates serious theoretical and logical weaknesses:

• Even if an RPM aide cannot move the arm of a person with a disability (as in FC), the aide can still move the letter board in such a way that the person with a disability appears to be pointing at any given letter. Thus, communication can still occur due to the aide’s actions, not those of the person with a disability.
• There’s also issue of prompt dependency. It’s been suggested that apparent communication from individuals with disabilities through RPM reflects an extreme form of prompt dependency, whereby individuals learn to respond to the prompt.  If this is true, then the communication would, again, originate from the prompting aide, not the individual with a disability.
• RPM presumes that any person with a disability has hidden knowledge and can comprehend spoken and written language, or at least be rapidly taught these skills. It does not simply presume that there are some people with disabilities who have hidden knowledge of language (which is clearly true), but it effectively presumes that all people with disabilities have such knowledge (which is almost certainly false).
• The rapid prompts used in RPM resemble the prompts and demands for compliance involved in Discrete Trial Training (DTT), a form of ABA intervention which has been widely criticized by autistic advocates.
• While the lack of current empirical support for RPM does not necessarily distinguish it from some other augmentative & alternative communication (AAC) techniques, the suggestion that RPM cannot or should not be tested is deeply disturbing. It is, in fact, eminently possible to test the efficacy of RPM.  FC was disproven when the facilitator and the person with a disability were independently presented with different information and it was discovered that the communication reflected the information available to the facilitator, not the information available to the individual with a disability.  Thus, the communication must have originated from the facilitator.  A similar procedure could easily be used to test RPM.  That those involved in developing and promoting RPM discourage such testing raises serious questions about their motives.

Thus, substantively, there may be little reason to assume that RPM is any better than FC.

It is important to recall that the mere existence of cases in which RPM has been used effectively does not prove the efficacy of the technique, and the same is true of FC.  I am a speaking autistic, and if I were placed in front of a letter board, I could easily point to the correct letters.  In that sense, RPM could be said to “work.”  The question is not whether there are some people who could communicate with RPM, since anyone who comprehends language and who has enough motor ability to point at a letter board could communicate with RPM.  The real question is whether the technique works for those people with disabilities who actually use it – for each of those individuals.

The Ethics of Testing Communication

As Sue Fletcher-Watson reminds us in her blog post on this subject, a demand that RPM should be tested is not about denying or questioning the right of a person with a disability to communicate.  On the contrary, the demand is that the RPM aide should be subjected to testing.  This testing is not aimed at denying communication to a person with a disability, but determining whether the RPM aide is in fact denying the person with a disability the right to communicate by falsely communicating on their behalf.

Fletcher-Watson warns that, in light of the danger that the RPM aide may (consciously or unconsciously) be communicating instead of the individual, “RPM as a communication support becomes less a tool for autonomy of the non-speaking autistic individual, and more a method of exploitation and potentially abuse.”  If an aide is violating the dignity and autonomy of a person with a disability by communicating instead of them, and presenting this communication as the individual’s own, this is arguably a far greater breach of the individual’s right to communicate than if the individual were simply denied communication.

Furthermore, we know that it is possible to test RPM.  We we can present information to the aide and to the individual with a disability, and we can do so in such a way that neither sees what is shown to the other.  Sometimes we can give them the same information; sometimes we can give them different information.  If the answers to our questions reflect the information given to the aide, and not the information given to the individual, this would show that RPM does not work for the pairing of the aide and the individual in question.

But Wait – Isn’t Individual Testing the Point?

As noted earlier, the current position of ASAN may not be that RPM should always be assumed to reflect genuine communication.  Instead, ASAN’s open letter argues that “each non-speaking person has a right to use the method of communication that works best for them, as determined by an individualized analysis.”  Doesn’t that mean that we would test whether RPM works using a procedure such as the one outlined above, thereby adequately protecting people with disabilities?

(“Individualized analysis” is a somewhat vague term, but one could certainly hope that a rigorous efficacy test was the intention behind the statement.)

Furthermore, we’ve also established that RPM could, in theory, work for anyone who comprehends language and who has some basic motor skills.  Doesn’t this make ASHA’s proposed position – that we should never, ever believe RPM reflects genuine communication – far too crude?

Well, perhaps it is crude.  But there are also good reasons to adopt a cautious attitude towards RPM.  If we demand proof that RPM is genuine communication in each individual case, how much time is enough before we can test the approach?  If someone is just starting with RPM, how long until we can test them?  It would be pretty unfair to expect instantaneous results.  But if we give the RPM aide a long enough time to try their best, we risk denying genuine communication and violating the basic rights and dignity of the individual with a disability by subjecting them to the experience of having another speak falsely on their behalf.

There’s also the issue of whether families and RPM aides would actually consent to genuine testing.  Recall that RPM people discourage such testing.

More importantly, the logic of RPM is really bizarre.  Why would anyone refuse testing?  And why is the aide’s holding of the letter board a necessary step in the RPM procedure?  Many people communicate with letter boards, and as long as nobody holds the letter board, there can be no doubt that their communication is genuine.

[Author’s note: In retrospect, I’m not sure what I meant here by criticizing the holding of the letter board.  I realize the facilitator holding the board may be necessary to provide some motor stabilization, or to try and grab the person’s attention before their attention gets captured by some other stimulus in their environment.  I may have been thinking that there would be other ways of providing motor stabilization, but I guess the human contact in particular might be helpful.  That being said, I think the important point here is that refusing testing is problematic.  If it works, why refuse to test it? Added Dec 12, 2019.]

So, How to Proceed?

I think we’ve established that there is good reason to believe that RPM probably doesn’t work for many people.  It’s certainly clear that there is great potential for abuse in the technique.  It’s clear that there is a considerable danger that the aide will begin to communicate instead of the individual.  If we are trying out different forms of AAC for a non-speaking individual with a disability, we could certainly try out a letter board and some form of motor stabilization, but RPM should not be tried.  We have no reason to recommend an approach like RPM over the other AAC techniques available.

The position of those individuals with disabilities already using RPM is more complicated.  As noted earlier, an individual who comprehends language and who has even basic control of their motor functions should be able to use RPM.  It is also possible, as ASAN suggests, that some of these individuals may prefer to use RPM.[2]  Thus, a decision to never recognize RPM as valid communication, as suggested by ASHA, could result in some individuals losing access to a communication method that does work for them.  Such cases could be “grandfathered.”  These individuals have the right to continue to use RPM if we know the technique works for them, and that their preference is accurate.  If an individual with a disability communicates through other techniques besides RPM and FC, they could use any these techniques to state that they would like to continue using RPM.  On the other hand, if an individual only communicates through suspect techniques like RPM and FC, their RPM aide would be required to successfully pass a rigorous test comparable to the procedure discussed earlier.

Personally, given the approach’s many flaws, I doubt there is much genuine need for RPM (as opposed to other communication methods more complex than the typical simplistic AAC forms) in the community, and the ASHA proposal to simply dismiss the validity of RPM outright does seem like a simple way of protecting individuals with disabilities from the danger of having their communication hijacked by an aide, the danger of having their right to communicate stripped from them and having someone else’s words put into their mouths.  However, this simple solution could violate the autonomy of some people with disabilities who already use RPM and prefer it.  If we all are careful to acknowledge the many flaws of the technique, and the immense dangers inherent to it, and if we all accept the necessity of robust procedures to avoid these dangers, then there does not appear to be any reason why we cannot respect the preferences of these individuals.

But to deny the flaws of RPM, to deny the risks and dangers inherent to it, could place non-speaking people with disabilities at risk of exploitation, of having their dignity and right to communicate violated, and perhaps even at risk of abuse.

This is a controversial subject, and I welcome comments – critical or otherwise – as long as they refrain from personal attacks.

[Author Note: Further to my last note, if I had to rewrite this post today, I think I would be a little less critical of some aspects of RPM’s methodology like the prompting or the holding of the letter board.  I’m not saying these are necessarily the only ways of capturing the person’s attention, and I still think we should be creative in trying to find other communication methods that would be richer and more natural than conventional AAC systems (for I certainly agree that if someone understands language we need to identify this and make sure their right to communicate is respected – I’m hoping we might be able to use something like eye-tracking or event-related potentials to index language comprehension in a passive assessment with no task demands), but I can see why RPM was designed the way it was and I can see how it might help a subset of individuals who struggle both with motor coordination and with their attention constantly wandering or being captured by other stimuli.  Maybe we could use letter boards designed with a particular size and angle to make them easy to reach and salient in attention, perhaps along with prompts as needed?  If we continue using RPM, the problem of testing will remain, at least in cases where the individual doesn’t immediately graduate to typing independently.  Although RPM adherents seem reluctant to test their approach, it seems eminently testable to me.  One merely has to control all factors except the facilitator’s knowledge of what is to be communicated: one must sometimes give the same information to both the RPM facilitator and the autistic person, and sometimes the RPM facilitator must be given different information.  Perhaps communication is only sensible when the facilitator has the same information as the autistic individual, in which case we could conclude that the facilitator was responsible, not the individual.  Perhaps communication is sensible even when the facilitator doesn’t have the same information, in which case we could safely conclude that the individual communicated.  Alternatively, perhaps communication is sensible in neither case, in which case the context or content of the test might have been inappropriate – perhaps the individual was stressed, or anxious, or having trouble concentrating in that context – in which case we can try to modify that testing situation.  It concerns me that RPM advocates resist empirical testing procedures like these that could be used to validate their claims that communication is occurring. Added Dec. 12, 2019.]

Footnotes

[1] For example, there is a passage in ASAN’s open letter characterizing ASHA’s proposed recommendations against RPM as being “based entirely on the absence of message-passing studies supporting the specific method in question.”

[2] Personally, I don’t see why someone would want an intrusive prompter holding a letter-board rather than a plain old letter-board, but it’s not my judgement that counts here.