I’ve decided to write this in order to essentially revisit a topic that I last addressed in a post a couple of years ago.  This post, regarding the Rapid Prompting Method (RPM), has (based on number of views) apparently received a lot of traffic compared to other posts on this blog, no doubt because of the vital importance of protecting people’s right to communicate.

Being profoundly aware of the importance of the right to communicate, as well as fully cognizant of my own relative lack of expertise in this specific area,¹ I’ve been reflecting on what I originally wrote. I’m not satisfied with what I wrote before and have decided to write a new post about RPM. Because it seems like the debate around RPM is closely linked to the shortcomings of Facilitated Communication (FC), I also thought I should cover FC as well.²

I’ve organized this new post around a series of statements regarding RPM and FC that I hope will strike most people as reasonable. 

What are RPM and FC?

Before going further, I should briefly define FC and RPM for anyone who is unfamiliar with these techniques.

First, FC: adherents of FC claim that at least some nonspeaking autistic people can type if their hand is held by a facilitator, who thereby grounds the individual and offers motor stability.  Clearly, this technique carries with it a strong danger that the facilitator can (perhaps unconsciously) start moving around the autistic person’s hand and controlling what they type, and research suggests this is usually what is happening in FC.

RPM is newer.  While practitioners in RPM (and FC, for that matter) may ultimately aim for the individual to graduate to independent communication, before this independence is achieved, the RPM practitioner will physically hold up a small number of response options, and later a board with letters that an individual could point to.  While the RPM practitioner is holding the individual’s response options, and prompting them to answer, there would seem to be a similar danger that the RPM practitioner may end up controlling the communication.  For this reason, some people are seriously concerned that RPM could actually be suppressing nonspeaking people’s right to communicate.  However, there are cases in which RPM users have graduated to communicating independently, and some people fear that any measures to restrict access to RPM might end up silencing nonspeaking autistic people who could communicate with RPM.

The Right to Communicate

Communication should be regarded as a basic human right.  I hope this particular premise should be self-evident to everyone.  We frequently talk about freedom of opinion and expression, but the right to communicate is even more fundamental.  It is not just about one’s political opinions, but one’s right to be able to communicate essential thoughts and needs.  Clearly, we must take all reasonable measures to ensure that people’s right to communicate is respected.

The Limitations of PECS-like AAC

PECS and similar AAC systems, including both low- and high-tech variants, are somewhat limited in the complexity of the communication they permit.  When I speak of PECS-like AAC, I am thinking of systems wherein individuals can communicate by pointing to or otherwise selecting pictographic icons (accompanied by words) that depict concepts.  While these systems are likely appropriate for some individuals, they are limited in the options for communication they present.  These options are selected by professionals and parents, who might create icons that address certain concrete needs (e.g., that the individual wants a particular food), but these selections might not allow for communication of any needs that aren’t anticipated by others, let alone more abstract thoughts.

I am not, of course, denying the utility of these approaches when used for the right individuals.  These systems can allow individuals who have a limited comprehension of language a level of communication that would not otherwise be possible, which is clearly highly desirable.

Furthermore, for some, these relatively basic forms of AAC might be a springboard towards richer forms of communication.  Spoken languages are obvious examples of richer forms of communication, but there are other alternatives.  Sign languages are certainly one important option.  Typing is another.  However, all of these methods often entail complex motor challenges, as well as attentional demands.  If there are people who have severe motor challenges but who have an intact understanding of receptive language and the desire to communicate complex ideas, they may need alternative forms of communication.

The Existence of Able Nonspeaking People

There are some people who have the basic cognitive capacities to communicate fluently, but who do not have enough control over motor function and attention to make their voices readily heard.  Admittedly, science knows fairly little about these individuals, at least within the autism world.  Our evidence regarding them is mostly anecdotal.  However, there are clearly people who have been diagnosed with autism and who have proven (against professional skepticism) that they have the ability to communicate despite motor difficulties that seriously delayed or prevented the development of spoken language.  Individuals like Naoki Higashida, Ido Kedar, and Tito Mukhopadhyay have shown themselves to be capable of fluent, eloquent, and independent use of language through typing and letter boards.  To be sure, these remarkable people didn’t immediately jump to being able to type, but after much practice, they did eventually graduate to typing independently.

We also have fascinating new quantitative evidence of language understanding in a sample of nine nonspeaking or minimally-verbal autistic young adults who appeared to be successfully using RPM or at least something very similar (the study doesn’t name the technique) to communicate. Crucially, these nine individuals had previously received an average of 13.5 years of conventional speech therapy and conventional AAC, but this was unsuccessful – it did not lead to effective, fluent communication. I’ll be speaking more about this study, which was conducted by Vikram Jaswal and colleagues, later.

The Need for More Research on Language Comprehension in Nonspeaking Autistics

We need better methods of detecting and measuring language comprehension in nonspeaking autistics.  A few anecdotal cases like Higashida, Kedar, and Mukhopadhyay might be enough to convince us that it is possible for language comprehension in autism to be hidden, but these anecdotes don’t answer the critical question of whether and to what extent any particular nonspeaking autistic person might understand language.  I suspect there’s a great deal of variability and heterogeneity in language comprehension in nonspeaking autism, but we don’t really know.  Traditional assessments of language understanding make assumptions that individuals have reasonably good motor function, have reasonably stable attention, and are interested in responding to the social demands of the adult administering the test.  We cannot necessarily make these assumptions with all autistic people.  In order to direct people towards communication techniques appropriate for their level of understanding, we need measures that can discern individual differences of language comprehension without the need for an overt, intentional behavioural response.  Unfortunately, this line of research is in its infancy.

One approach might be to use event-related potentials (ERPs), or the electrical response emitted by certain neurons in the brain’s cerebral cortex as averaged over many presentations of some stimulus.  There are ERP responses that seem to reflect semantic understanding – the understanding of the meaning of language – in neurotypical individuals.  Several studies have explored whether these responses can be observed in nonspeaking autistic individuals (e.g., Cantiani et al., 2016; Coderre et al., 2019; DiStefano et al., 2019).  However, their results have been fairly inconsistent.  Indeed, in one study, a late negative ERP response yielded significant evidence of semantic processing in nonspeaking autistics, but this effect did not achieve significance in speaking autistics (DiStefano et al., 2019)!  As a researcher who uses ERPs to study individual differences rather than just group differences, I do believe that ERPs can carry meaningful information about individuals, but I’m also aware that there are many factors that can affect an individual’s ERPs.  Therefore, it may be problematic to draw firm clinical conclusions that could affect an individual’s life solely from ERP evidence of semantic processing, although this technique has the advantage of being purely neural – there is no need for the individual to make any sort of motor response.³

Another option might be to use eye-tracking to index receptive language in autism.  One could display pictures of something and something else, then measure whether people look more at whichever of the two things is verbally named.  This approach has been used in autism by various researchers (e.g., Brady et al., 2014; Coderre et al., 2019; Venker et al., 2016, 2019), including by my mentor Susan Rivera, who introduced the task to me.⁴ Our autism results haven’t been published outside of posters just yet, but we’ve found robust correlations between eye-tracking indices of language understanding and traditional measures in autism, suggesting these tasks are validly measuring language, but hopefully with fewer task demands.  A published study from the lab found similarly robust convergent validity with other language measures in Fragile X syndrome (Yoo et al., 2017).  However, just as I don’t think ERPs alone are perfect, I would not suggest that this eye-tracking paradigm is clinically reliable.  A few people might lack oculomotor control.  Others might understand words but not automatically orient attention towards those words.  There’s also a lot of noise in the data, so many trials might be necessary to get a good average measure of visual preference.

However, the previously-mentioned Jaswal et al. study of non-speaking and minimally-verbal young adults used a different eye-tracking method that is more directly relevant to questions around RPM. In this study, participants wore a head-mounted eye-tracker, sort of like a hat that both tracked what was in front of participants and where specifically their eyes were looking. Jaswal et al. found that participants looked directly at letters before they touched them, which suggests the participants knew what they were doing: the RPM professional wasn’t directing them to point at a particular letter. The participants also responded quickly and accurately. There were even delays around word boundaries that suggest the participants understood where words ended and began. This study seems to offer pretty convincing evidence of language understanding at the group level, and perhaps even the individual level. Now, admittedly, participants appear to have been deliberately selected because they were “considered by staff members at the centre [where the letterboard communication method was taught] to be skilled letterboard users,” so the findings may not be representative of all nonspeaking autistics, but they at least show the possibility that language understanding and typing production can be measured this way.

Unfortunately, the Jaswal et al. method seems to have one major limitation: the researchers had to manually code tens of thousands of video frames from each participant, indicating where they were looking at what they were doing at any particular moment. I’m sure it took them a long time and a lot of effort! Until technology advances, this may limit our ability to apply the method widely. Their method also assumes that people have already started learning to communicate with a letterboard, and it does involve language production as well as comprehension.

Still, we clearly have some potentially useful implicit measures of receptive language in autism.  Eye-tracking, in particular, is looking very promising.  Because of the critical importance of developing new ways of looking at language understanding in autism, it seems like we should work on further improving these measures, developing more measures, and (ultimately) getting them packaged in such a way that they can be taken out of the research laboratory and into the real world.

The Need for Alternatives to The Usual Autism AAC

When we identify people who can understand language, but who cannot communicate at a level of fluency commensurate to their understanding using spoken language or traditional AAC, we need to be able to offer them alternative ways of making their voices heard. These are people for whom spoken language, sign language, and even typing are (at least initially) too demanding.  PECS-like approaches might be useful for some, but even these might be problematic for some individuals with severe motor or attention problems.  Other people might be able to use PECS-like AAC, at least to some extent, but their true level of understanding might far outstrip the limited capabilities of PECS and similar systems.  Thus, we need other options.  Could FC and RPM offer these options?

The Need to Verify that Communication is Genuine

Protecting an individual’s right to communicate means not only offering individuals access to communication systems that we think are appropriate to their needs, but also verifying that the communication systems are actually appropriate and that they are actually allowing for successful communication.  If a practitioner ends up taking over the communication, they have denied the individual their right to communicate.  Such a violation of the right to communicate would be unacceptable.  When we try to empirically verify that any given individual who is working with an FC or RPM practitioner is actually driving the communication, this should not be framed as being about testing the autistic individual or demanding that they prove themselves.  On the contrary, it would instead be about testing the practitioner to make sure they are not suppressing the autistic person’s communication.

I feel like a concrete example may be helpful here.  In the 1990s, Janyce Boynton tried to use FC with a nonspeaking autistic child, Betsy Wheaton.  After a while, Boynton and Wheaton began typing out increasingly graphic accusations that Wheaton was being abused at home.  However, after a while, empirical testing demonstrated quite conclusively that Boynton, not Wheaton, was driving the communication.  Looking back, Boynton (2012) realized that Wheaton’s behaviour had actually communicated that she wanted the FC to stop: Wheaton had begun to respond with frustration and aggression when Boynton intruded into Wheaton’s space in order to hold her hands and engage in FC.  However, at the time, Boynton still believed in FC: she didn’t want to think that Wheaton was objecting to the FC.  Instead, Boynton decided that Wheaton must be acting out because she was a victim of abuse, and since Boynton controlled the communication, references to abuse began appearing up in the FC sessions.  Thus, in this case FC not only led to spurious communications and specious allegations that undoubtedly caused great distress to Wheaton and her family (Wheaton was temporarily removed from her home by social services), but the FC led Boynton to ignore what she describes as “the clearest communications” she ever received from Wheaton through her behaviour: Wheaton’s nonverbal protests that showed she wanted the FC to stop.

Nor is this the only concrete example. Consider: what if FC was used to manufacture a false consent to sex? This is not a theoretical issue; facilitators have, in at least two cases, sexually assaulted clients following false sexual consents obtained through FC.

Therefore, if we intend to use techniques like RPM and FC with individuals who cannot communicate independently to verify that their words and thoughts are being accurately transmitted, it is imperative that we make certain that the techniques are working.  If the technique does not work – if the practitioner communicates instead of the autistic individual – then the autistic individual’s fundamental right to communication has been violated.

We mustn’t deny somebody the right to communicate by preventing them from accessing a communication method that could work, but nor should we suppress their communication by using a method where the practitioner produces spurious communications.

It is Possible to Test the Accuracy of Communications in RPM and FC

We have the means to empirically determine whether someone who cannot communicate independently is successfully communicating using RPM or FC.  Advocates of RPM and FC have sometimes objected to testing on the grounds that the testing situation may cause anxiety and stress, which in turn could reduce the quality of the individual’s communication.  While some nonspeaking autistics do indeed report such anxieties and stress that can interfere with motor and attention control (e.g., Kedar, 2012, pp. 58-59; Mukhopadhyay, 2008, pp. 115-116), I believe there are ways that we can design the testing situation so that any such anxiety or stress will not compromise our conclusions.  Of course, we can and should do everything in our power to treat people with respect and to design a testing environment that will be as stress-free as reasonably possible.  For example, the Jaswal et al. study I keep bringing up seems like a perfect example of a method designed to dramatically limit demands and stress. However, unfortunately that particular method seems to depend heavily on laborious, time-consuming manual coding that may limit its use with individuals in community settings outside the research laboratory. Still, if we can’t use a method that will entirely alleviate the individual’s stress and anxiety, we can at least do the best we can and reassure ourselves that the damage caused by such brief discomfort would likely be outweighed by the damage caused by stripping away the individual’s right to communicate by allowing another to control their communication.

One basic way of going about testing is to set things up so that the autistic individual and the facilitator are both presented with information: sometimes the same information, and sometimes different information.  In an FC context, this can be accomplished by, for example, erecting a screen between the autistic individual and their facilitator, who would sit side-by-side.⁵ Both would look ahead and would be presented with information on a screen, folder, or display, but neither could look and see what the other saw.⁶ This is the approach of Wheeler et al. (1992) and various other studies, and it can allow for clear conclusions:

• If the autistic individual can communicate in a sensible way (i.e., in a way that relates to the information presented to them) in all conditions, they can clearly communicate using the approach.  The facilitator has not caused the communication to occur.
• If the communication is only sensible in the condition where the facilitator and the autistic individual both see the same information, and not in any other condition, the facilitator has caused the communication to occur.  The communication approach is suppressing the autistic person’s own communication.⁷
• If the autistic individual cannot communicate sensibly in any condition (despite being able to do so in daily life), there are at least two possibilities.  First, the autistic individual might genuinely have been able to communicate, but they might have become anxious and this could have detrimentally affected their performance.  Alternatively, the facilitator might have been driving the communication, but they might have become anxious (perhaps subconsciously realizing that they were driving the communication),⁸ leading to garbled answers.

In the first and second cases, we’re done.  We have the answers we need.

In the third case, we may need to try something else to figure out which of the different possible explanations is the right one.  Perhaps we could have another look at the testing conditions and see if there is anything else we can do to minimize anxiety, or perhaps we could try another test paradigm.  One such paradigm is that used by Roane et al. (2019), who evaluated whether a nonspeaking autistic individual could behaviourally display knowledge of words (taken from FC transcripts, so supposedly known to the individual) by, for example, pointing to them when they were indicated.⁹ As a control, Roane et al. also presented the autistic person with pictures of things and asked the person to point to the corresponding picture from an array.  A verbal receptive language task was also administered.  In the case study by Roane et al., the individual was able to match pictures and to show verbal receptive understanding, but they were not able to demonstrate knowledge of the written words they had supposedly produced in the FC sessions.  From the intact verbal receptive language and picture understanding, we can conclude that the individual was motivated to engage in the task, but the lack of written comprehension suggests the facilitator was responsible for communication in the FC sessions.^{10 11}

Granted, we would need to give an individual some time to learn in an AAC approach before we start testing – for example, in the Jaswal et al. study, individuals had been learning for a minimum of two years. This begs the question of when an evaluation should be done. This is a complex question, but if someone does appear to have learned successfully, it seems like it would be appropriate to empirically verify that they are responsible for their own communication.  By the time a client is appearing to communicate competently, one would hope that their response would not depend on the desired answer being embedded within an RPM practitioner’s prompt, but that they would be thinking independently and supplying their own answers.

Therefore, there do seem to be ways by which we can empirically evaluate whether individuals who communicate non-independently are responsible for their own communication, or whether the communication is produced by the practitioner.  It seems that these tests should be employed for individuals using non-independent AAC methods, in order to ensure their right to communicate is being respected.

Of course, if somebody moves on from FC or RPM to typing or otherwise communicating independently, we can assume that they would tell us if the FC/RPM-assisted communications are wrong, so this would eliminate any need for testing!

FC is Vulnerable to Spurious Communications

While it is theoretically possible that FC could work, repeated empirical testing suggests that it is seldom effective in practice: instead, it seems as though communication is generally driven by the facilitator.  I say that FC could work in theory because a person who does understand language could use FC.  For example, I’m a speaking autistic person and I clearly understand language.  If I asked someone to hold my hand while I typed, I would still be able to type fluently.  The question of importance is whether this is actually occurring when FC is used with most nonspeaking autistics.  Regrettably, although there may be occasional successes – for example, Naoki Higashida’s mother seems to have supported him using a method similar to FC before he began communicating independently; Sue Rubin now communicates independently after initially learning to communicate with FC; and before Ido Kedar began using RPM, Kedar’s mother first realized that the extent of his language understanding when he began to write while she held his hand, etc. – empirical research investigating FC using the methods I described above is quite clear: in practice, it seems like facilitators usually drive the communication (e.g., Wheeler et al., 1992; Montee et al., 1995).

Of course, I suggested that these tests should be used at the individual level, because whether a technique might work could vary from individual to individual.  That said, there seems to be clear evidence that FC doesn’t work in practice for the majority of individuals who use it.  Given this, is FC really an approach we should try to use?  Again, I’m not an expert in this area, but I have to say it certainly does not look like an attractive first option.

RPM Practitioners Could Do More To Prove their Method

While RPM can sometimes work, RPM practitioners certainly don’t seem to have proven that their method works for all of their clients.

Of course, specific individuals have learned to type independently after learning through the RPM method – Ido Kedar and Tito Mukhopadhyay are both examples.  Given Kedar and Mukhopadhyay’s experiences, and Kedar’s eloquent advocacy for RPM, we certainly can’t dismiss the method outright! Furthermore, Jaswal et al. seem to have shown quantitatively that RPM can work for a number of nonspeaking and minimally-verbal autistic individuals.

I should also say that there are things about the RPM method that make a great deal of theoretical sense.  If the individual’s attention is constantly wandering or being captured by things in the environment, rapid prompting might indeed be helpful in focusing attention.¹² If an individual has motor difficulties, holding a letter board where it can be easily seen and reached might be helpful.¹³

However, it’s also hard to deny that there are clear vulnerabilities in the RPM approach: for example, the RPM practitioner’s prompting might have unintentional demand characteristics that could encourage someone to pick one response option over another.  The practitioner might unintentionally move a response option or the letter board so that the autistic individual makes a particular response selection.¹⁴

Of course, we’ve already established that I think any claims of success at the individual level should be tested – we need to make sure RPM practitioners are doing their jobs and that the autistic people are the ones doing the communicating, not the RPM practitioners.  Ultimately, the fact that there isn’t a clear empirical evidence base demonstrating that RPM practitioners don’t hijack autistic people’s communication (see Schlosser et al., 2019) makes me somewhat suspicious. The Jaswal et al. study goes a long way towards filling this gap, but if its participants were specifically selected for being outstanding communicators, can we assume the findings generalize to all RPM clients?

It’s true that we do some things without empirical evidence and on the basis of subjective feelings and common sense – in fact, I’m becoming somewhat critical of what I see as an excessive emphasis on evidence-based practice in some circles – so I must concede that simply saying there’s no evidence for something isn’t necessarily a decisive blow against that thing.  However, given that non-independent communication in FC can be hijacked to deny the individual opportunities for communication, it seems like we have special reason to worry about anything involving non-independent communication, and that would include RPM before people graduate to independent typing.  Moreover, as noted earlier, communication is a basic and fundamental right; if a facilitator was falsely generating communication, this would be a violation of this right. I think this is reason enough for us to test RPM and FC more rigorously than we would test other AAC techniques.  If RPM practitioners think they are helping their clients and not hijacking their communications, then surely they have an obligation to prove this through better individualized assessment of clients?

I hope that RPM is being used successfully and responsibly with all RPM clients, but I’m not sure we know for sure either way yet.  I hope that RPM practitioners will begin having their performance tested and that the testing will prove that any fears about RPM are unfounded, but currently it seems like testing and assessment are not occurring in communities.

But Are the Usual Arguments Against RPM Helpful?

While we’ve seen that there seem to serious potential issues with RPM – and that RPM practitioners have some obligation to lay our fears on that score to rest, opponents of RPM sometimes seem to rely on inflexible, dogmatic attacks that probably aren’t terribly helpful.  For example, I sometimes see attacks on RPM that simply assert that there is no evidence base to support RPM, that no intervention should be employed without evidence, and therefore that RPM should not be employed.  However, as noted earlier, an over-reliance on empirical evidence has its limitations.  Granted, RPM does seem to be a special case where clear quantitative empirical evidence might be quite necessary in order to protect individuals from having their communication suppressed.  But if the attack on RPM is solely based on a lack of evidence, period, then adherents of RPM can quite correctly rebut the attack by pointing out that not all interventions that we believe can work well – indeed, not even all conventional AAC techniques – rest on a thorough quantitative evidence base.

Another issue with this simplistic, dogmatic sort of attack is that it ignores the existence of powerful anecdotal evidence in the form of successes by Kedar and Mukhopadhyay. The study with the Jaswal et al. sample is newer, but it is an equally important source of evidence.  Now, these individuals might very well turn out to be unusual exceptions, but it seems like some opponents of RPM would prefer to ignore their existence by asserting that there is no evidence at all to support RPM.  The independent communication achieved by Kedar and Mukhopadhyay suggests that the picture is more complicated than this: RPM might or might not work for all nonspeaking autistics, but clearly it works for some. Unfortunately, instead some researchers have argued that Kedar’s words should be ignored due to his use of RPM, in spite of evidence that Kedar can communicate independently as well.

Attacks on RPM, then, can sometimes lack some of the subtlety or flexibility that might be demanded by the complexity of the issues at hand.  This is clearly problematic, even if the opponents of RPM do turn out to be partially right.  Inflexible and dogmatic attacks on RPM are presumably relatively easy for RPM adherents to refute.  Ironically, then, these attacks could actually make it easier for RPM adherents to dismiss demands for greater accountability for RPM practitioners.  Moreover, dogmatic arguments might heighten community distrust in the research and professional “establishment,” as it were.  If parents who desperately want to help their nonspeaking children by any possible means observe that researchers and self-proclaimed evidence-based practitioners are willing to simply ignore evidence of successes achieved by RPM, why would these parents take evidence-based arguments seriously?

1. Full disclaimer: I am a speaking autistic person who has never used augmentative and alternative communication (AAC).  Furthermore, none of my research is about AAC.  Thus, I cannot make claims to possessing either experiential or academic authority or expertise in the area of AAC.
2. This is NOT an effort to suggest that RPM and FC are the same thing; on the contrary, I would say that in general I’m quite a bit more skeptical of FC than RPM.
3. In fact, you want the individual to avoid any motor movement that will produce artefacts in the EEG.
4. This is the one and only part of this particular blog post where I can claim to have some experience relevant to the topic I’m addressing.
5. I understand RPM is also supposed to happen side-by-side, so I guess this testing method could be transferred to an RPM context? But I’m no expert!
6. Furthermore, one can add a condition where only the autistic person is presented with information and the facilitator sees nothing.  If there is less communication in this condition, it may be because the facilitator is unable to guess what the autistic person might have seen and is therefore falling silent out of uncertainty.
7. In this case, we could be confident that the autistic person’s anxiety was not a factor, because the autistic person wouldn’t know when the facilitator sees the same information or when the facilitator sees different information.  Thus, they wouldn’t be any more or less anxious in either condition.
8. Assuming, of course, that the facilitator is not consciously aware of controlling the communication and trying to avoid being found out.  But I suspect (and hope) that such intentional malevolence is very rare.
9. We might also try to use eye-tracking to explore this, eliminating the need for an overt behavioural motor response.
10. In the specific case of the study by Roane et al., the picture, receptive language, and written language tasks were not controlled so as to be precisely equivalent apart from the specific manipulation that different stimuli were used in the different conditions; the conditions differed in many other ways as well.  In any future application of this testing approach, such differences between these tasks should probably be eliminated, so as to remove any potential confound.
11. This is particularly unfortunate because the verbal receptive language comprehension of this individual suggests a different form of AAC might have allowed them to successfully communicate.
12. That said, for another individual, it seems possible that rapid prompting might be experienced as intrusive and a source of sensory distress.  This is another reason why learning more about individuals is so important!
13. Indeed, to make a more general point, it seems like our concept of autism has shifted over time. To oversimplify a little, we used to think the basic experience of the world was the same in autism and TD, and that autistics just differed in some kind of social processing or social attention. That may be true for some of us, but it is increasingly clear for many that differences in our basic attentional processing, sensory experiences, and motor abilities are central. Different interventions would seem to make more or less sense depending on which of these models holds for a particular individual.
14. RPM practitioners may also sometimes touch children (e.g., Ochs et al., 2005), which of course raises questions similar to those in FC, though it is not clear to me how extensive the problem is in RPM.