Tag: pragmatic

Posted on

Using AI-Generated Music to Teach Social Scripts

I have to say from the start that this is not a paid endorsement, and endorsing specific products is not something I normally do, but I’ve discovered a tool that has been a game-changer for how I work with one of my direct services clients as his counselor. That solution is using AI to create songs that are individualized to a specific person with lyrics based on therapeutically appropriate social scripts tailored to the person.

I discovered this solution in the course of looking for ways to embed peer-reviewed music therapy elements into my counseling sessions with this client because he is highly responsive to music, and seems to remember lyrics set to music better than spoken words. Given that he’s lost his eyesight, we’ve got to rely on his other senses.

I was looking for an easy way to generate songs he would take seriously as legitimate musical productions that contained the social scripts, such as “safe hands,” “inside voice,” and “be patient,” with which he was already familiar in order to expand his understanding and application of these concepts in his day-to-day life. I’m still shocked at how easy it was with the AI.

These individualized songs are also serving as a stepping-off point to teach my client new, more sophisticated social/emotional skills and scripts, going forward, once he’s incorporated them into his music listening routines and we work with them in our sessions. As time goes on, I’ll be adding new songs that tackle more sophisticated concerns than those that I’ve initially created to get him started.

My counseling client is in his late 20s and lives in a group home with 2 other men who have developmental disabilities. My client is blind, autistic, and intellectually disabled. He struggles to produce spontaneous speech and relies largely on scripted speech to communicate verbally with others.

Since 2010, I’ve been this young man’s lay advocate, his attorney’s paralegal, his compensatory education services provider. He and I have gotten to know each other well and have instant rapport with each other, even after not seeing each other in person for a couple of years.

Given the friendship and rapport I share with my client, I guess I shouldn’t have been as surprised as I was at how quickly he took to the songs I created for him using AI, but I was actually flabbergasted. It was during my last session with him, in which I was collecting the last of the baseline data I needed to inform my program goals for him over the next 10 months, when I introduced the songs to him.

The moment I started playing the songs, the stimming decreased to nearly none and he sat listening, turning his head so his ears faced the music, and orienting to me as if looking me in the face to repeat familiar scripts he was hearing in the lyrics with a grin on his face. He was fully engaged and it took next to no effort from me. I was floored. I was sure that I was going to have to work to sell him on the idea, but he took to it like a fish to water.

This has left me inspired, because I know he can’t be the only one who would benefit from this. I was in an online IEP meeting for one of my other students a few days ago, and mentioned this experience to the other professionals who were already logged into the meeting, while we were waiting for the parent and a few other professionals to log in. When I told my colleagues about what I’d done using the AI with social scripts to create highly individualized music for therapeutic purposes, they got all excited about it.

So, based on the feedback I’ve gotten so far, I’m stopping what I’m doing right now to bust out this short post/podcast to share this information with anyone else who might benefit from it so that I can let it go and move on with the rest of my day. This is going to keep bugging me until I share it, and it’s preventing me from finishing anything else until it’s off my plate. Call me perseverative if you want; it is what it is. Thankfully, this can be fairly brief.

The music-generating AI website I stumbled upon after 30 whole seconds of Googling is MakeBestMusic (https://makebestmusic.com/app/create-music). Again, this is not a paid endorsement.

I didn’t compare this AI against any other. It was the first one I tried and it instantly gave me what I was looking for in just the free demo. I copied and pasted the list of social scripts that I wanted incorporated into a song, selected some genre-specific tags, and hit the “go” button, then a minute or so later, I had two new songs using the words I’d provided as lyrics and one of them was absolutely perfect. I repeated the process for three more sets of social scripts and ended up with a total of four songs.

For the sake of illustrating my point, I’m playing one of them, titled “Ask for Help,” here:

Listen to the lyrics and you’ll hear that they are clearly about social behaviors, but it’s sounds like a real song and not lame like something I’d make up if I had to do it myself. The robots do it better than me, and for these limited purposes, that’s okay.

I’m not trying to earn an award for high quality music. I’m trying to teach my client how to act right around other people and still live a happy life for himself. For those of us who could never afford to outsource this kind of work to a professional songwriter, AI is a sufficient tool for this type of job.

Given that a less than professional job can serve a valid therapeutic purpose using AI at a much lower cost, using AI-generated music to embed music therapy elements into a program of social/emotional counseling with individuals who have developmental disabilities can be an affordable and powerful tool in a counselor’s arsenal of solutions. I encourage my colleagues who work with individuals with needs similar to those of my counseling client to play around with this type of technology and see what kinds of solutions you can create.

If you’re the parent, you probably have even more ideas about how you could use this around your home with your own family members. Seriously think about setting the step-by-step instructions on how to perform certain chores to music to play when you have your kids helping you around the house. Once the song gets stuck in their heads, so are the instructions on how to perform those chores.

I think AI-generated music holds a lot of currently untapped potential for parenting, teaching, and therapeutic interventions, and I’m curious to see how other people use it in these kinds of ways as time goes on. What instructional, parental, and/or therapeutic outcomes can you pursue using AI-generated music?

Posted on

Pragmatic Language & YouTube Reaction Videos

Could YouTube reaction videos be used to teach pragmatic language skills?

I’m not a Speech-Language Pathologist (SLP), so I’m not pretending to be an expert in the field of language processing. However, I rely on data from SLPs to inform my understanding of the communicative aspects of individual learners’ respective abilities to process information and put it to constructive use.

I’m familiar enough with the concepts of language processing to have some informed questions about things I see in the world, every now and again. One of those things that just dawned on me most recently is the question of the relationship between pragmatic language processing and the popularity of reaction videos on YouTube.

For those of you who may be unfamiliar with reaction videos, they are videos made by YouTubers in which they react to videos that have become popular on YouTube, as evidenced by their respective number of views. So, to be clear, it’s videos of people watching videos, usually for the first time, so that other people can watch their reactions.

The pay-off of watching reaction videos is to connect with the reactor’s emotions through the reactor’s body language, facial expression, word choice, and tone of voice. Of those four elements of language watched for by the audience in a reactor during a reaction video, three of them are pragmatic language.

Here is my hypothesis, but I need the SLPs in our audience to weigh in on this, too: You know how when you see something cool, your first impulse is to share it with somebody else and see how they react to it? It’s like we only get one first time of experiencing something, but we want to relive it and the only way we can is to watch someone else experiencing it for the first time.

We ride the emotional roller coaster with each new first-timer we expose to the cool thing, relating to that other person’s emotional response based on our own memories of enjoying our first time with whatever the cool thing is. It sounds like a weaker version of the behavior we otherwise refer to as addiction. The first time is always the best time and the experience can never be fully recaptured, but it can be approximated. It goes to show that all behaviors occur on a spectrum, including those we typically regard as extreme.

Art is the manipulation of media in order to convey emotion. It is often non-linguistic. Light, color, sound, shape, space, and a host of other things can be manipulated according to the laws of physics to evoke feelings and tell stories without words. Other forms are art use words as one more medium to enrich their creations, whether written, spoken, and/or sung.

One of the most popular forms of reaction videos on YouTube is devoted to music, specifically individual music videos. This involves the manipulation of visual and auditory information, only, as the other three senses cannot be actively engaged. The exception could be bone conduction of vibrations from the music in reactors wearing headphones or near loud speakers, creating proprioceptive input that goes to the sense of touch.

There are dozens of reaction videos apiece to a great many songs on YouTube. The number of people reacting times the number of songs to which reactions can be given creates exponential exposure for the artist of each original performance video. Reactors increase their own exposure on YouTube by riding on the coattails of artists who have millions of views of their content because of the quality of their art.

When people search YouTube for an original artist’s work, all of the videos of people reacting to that artist’s work will also come up in the search results. It’s only natural that once one has viewed the original video to want to see it again through the eyes of someone else who has not seen it before and determine if they reached similar conclusions. People are not just looking to relive the experience, but also to be emotionally validated for feeling the ways they felt experiencing the original video for the first time.

Which then begs the questions, “Why do people get so sucked into these videos that are so heavily based on pragmatic language?” and “What are the implications of those facts for individuals who struggle with pragmatic language disorder or autism spectrum disorders that compromise their abilities to accurately read the facial expressions, body language, and tone of voice of others, and express themselves appropriately that way, themselves?”

This boils down to the research question of, “Can reaction videos be used to teach pragmatic language skills through video modeling to individuals who struggle with pragmatic language?” Only scientific research can tell. I’m all about encouraging such research, because now my inquiring mind wants to know.

One of the most powerful examples I can think of is the song, “My Mind,” performed live by Yebba at Sofar New York a few years ago. I have never heard anybody take people on such a hypnotic journey through sound in my life. Watching the reactors getting sucked into the song and becoming mesmerized is something to witness unto itself.

The impact of the reactions to her videos led to a compilation video of several reaction videos, that was basically the YouTube version of a meta-analysis, in which all of the reactors’ reactions were displayed simultaneously, allowing viewers to see which parts of the song triggered the strongest reactions from the most reactors at once, like a living performance graph. Me analyzing that now is like the reflection, within the reflection, within the reflection … like, a metaphorical nautilus of analysis.

Another mesmerizing performance is “SOS” by Dimash Qudaibergan at the Slavic Bazaar, also from just a few years ago. Watching people who have never heard of him before reacting to Dimash singing “SOS” is something to behold. The first time you watch it yourself, you’re immediate reaction is, “No! That can’t be real. He’s not human!” Then you watch it again in the reaction videos and see other people having their responses and you think, “Okay, it’s not just me.”

Another one that requires additional inquiry is Chris Stapleton’s “Tennessee Whiskey,” which doesn’t even have a video. It’s just the song with a still image of the album cover throughout, and yet it has over 500 million views on YouTube as of the time of this post. Watching people who have grown up on rap and hip-hop reacting to this song with surprise is a joy. They are the ones that give animated visual life to what is otherwise a largely auditory experience.

Anyone watching the Kodi Lee AGT audition reactions can see a handful of egocentric attention- and click-seekers suddenly reduced to puddles of humility over and over again. In an instant, Kodi’s performance puts things into perspective and they get it. The clicks to watch the reaction become earned because it isn’t a trick; these people are legitimately shook by what they see and that’s what engages viewers of reaction videos.

In all of the above-referenced original videos, surprise is always a key element. In every reaction video that gets any kind of traction on YouTube, the reactors are shocked by what they are watching for the first time, and become emotionally engaged with the song and performer to which they are reacting. In all the instances cited above, there is an emotional story being told with which listeners can identify.

The reason the views of the original videos are so high in the first place is because the content is so emotionally engaging. People reacting to them for the clicks suddenly forget about the clicks, find themselves transported, and start talking about things that actually matter in the world. What often started out as an exercise in narcissism for pay can become a transformative experience that snaps a selfishly motivated YouTuber right out of it and puts things into proper perspective.

The sounds of the originally performed songs conform with their respective story lines in a way that takes the listener along for the emotional ride of each. With the exception of the Chris Stapleton example, above, reactors also have the benefit of watching the performance, which adds the benefit of facial expression and body language to the communication. Each song conveys a different emotional experience, but one must have intact pragmatic language skills to appreciate what makes each song so uniquely impactful that it inspires so many views and, thus, so many reaction videos.

And, I want to be clear that, even if the reactors are initially reacting to these specific videos only for their own marketing purposes, the ones that get the most traffic are the ones in which the reactors are caught off guard and have authentic responses, like crying or, in the case of Yebba, getting moved by the Holy Spirit in the middle of a song that is not about religion in any kind of way. The value in watching these reaction videos is seeing real people moved for real in the moment without the opportunity to fake it.

There’s no way to conceal authentic surprise and awe, and those are the feelings viewers seem to be trying to experience by watching these reaction videos. What is it about the human psyche, then, that causes us to seek experiences that make us feel surprise and awe? Why do we want to witness miracles so badly? Why are the outliers who receive the most favorable public attention usually artists rather than scientists? Why do we tend to think data is boring and seek emotionally extreme experiences when data is practically useful and emotions often are not?

I don’t have the answers. I just think this is a line of inquiry worth exploring. I’m curious to see if the evidence in support of video modeling as an instructional strategy could be applied to using reaction videos to teach pragmatic language skills to those who struggle with this area of language processing. Are there any communication researchers out there who might want to conduct some studies so inquiring minds can know?

Posted on

California SLPs Sometimes Confuse Legal Requirements

Today’s posting will hopefully lay to rest a misunderstanding that seems to plague special education in California. I can only presume that, like many other “urban myths” that root themselves in special education lore, at some point in time, somebody somewhere in California conducted a training seminar on speech-language assessment and services within special education and miscommunicated something that has now led to speech-language specialists throughout the state making improper conclusions to the detriment of some children in need of speech-language services.

The problem is this: the distinction between who is found eligible for special education on the basis of a speech-language impairment (“SLI”) and who qualifies for speech-language services as a student already eligible for special education under any other category. Eligibility for special education as SLI is not required in order for a child otherwise eligible for special education to receive speech-language services in order to benefit from his/her IEP.

The critical piece of legislation, which gets erroneously cited in speech-language assessment reports all the time, is 5 CCR  3030(c). Title 5 of the California Code of Regulations Section 3030 describes all of the criteria for each of the eligibility categories under which a student may qualify for special education and related services. These categories include Specific Learning Disability (“SLD”), Other Health Impaired (“OHI”), Emotionally Disturbed (“ED”), and many others, including SLI. The critical thing to understand here is that the 3030s describe who can receive special education and on what basis, not what services they will get.

What ends up happening, though, is a child will be assessed for special education purposes and a speech-language assessment will be conducted. In the course of the overall assessment, even though the child is found eligible under some category other than SLI, because he did not score below the 7th percentile on two or more speech-language assessments, the speech-language specialist will determine that he doesn’t qualify for speech-language services according to 5 CCR   3030(c). It is a complete and utter misapplication of this Code, which deals strictly with eligibility under SLI and not what services an otherwise eligible child should receive.

A typical example of this would be a child who is eligible for special education pursuant to 5 CCR   3030(g) for autistic-like behaviors (in special education in California, a medical or psychological diagnosis cannot be made by the school psychologist, so this section of the code provides alternative language and defines the criteria by which a special education eligibility category can be identified for a child exhibiting the symptoms of autism), but who is relatively verbal. While his scores may hover just above the 7th percentile on the speech-language tests he was administered, they are still very low and his low language functioning compounds his other problems arising from the other needs arising from his handicapping condition.

In this example, anyone in their right mind can see that the child needs pragmatic (social) language intervention and help with idiomatic and figurative (non-literal) language. He doesn’t have any friends, he doesn’t get jokes, and he doesn’t understand clichs and colorful sayings, such as “Clear as mud.” This makes it difficult for him to participate in group projects with peers and understand the writings of Mark Twain. He needs goals that address these areas of need and speech-language services in order to benefit from his IEP.

No subsection of 5 CCR  3030 drives the selection of services that any child gets, only whether or not a particular child is eligible and, if so, under what category. The IDEA mandates that children who are eligible for special education, regardless of what category they qualify under, receive whatever supports and services are necessary in order to afford them a FAPE.

Specifically, the federal regulations found at 34 CFR  300.320(a)(2) state that IEPs must include for each child measurable annual goals, including academic and functional goals designed to meet the child’s needs that result from the child’s disability to enable the child to be involved in and make progress in the general education curriculum and meet each of the child’s other educational needs that results from the child’s disability.

An eligible child is a child who requires, as a result of one or more handicapping conditions, special education and related services in order to receive educational benefit. 34 CFR  300.39 “Related services” is described at 34 CFR  300.34. In none of this is there anything that suggests that the only way that an otherwise eligible child can receive speech-language services is if he is also found eligible as SLI.

In fact, 34 CFR  300.304(c)(6) states that, when evaluations are conducted for special education purposes, they must be “sufficiently comprehensive to identify all of the child’s special education and related services needs, whether or not commonly linked to the disability category in which the child has been classified.” Congress understood when it crafted the IDEA that you don’t individualize a child’s program by resorting to “cookie-cutter” strategies that are based on a kid’s eligibility category.

The IDEA is the skeleton of special education law. It establishes the basic framework and minimal standards. It is left to the states, if they want any federal special education dollars, to add the flesh to the bones by creating their own state-level legislation that explains how each state will implement the requirements of the IDEA. While states are free to add more obligations to their schools than what the IDEA requires, they are prohibited from reducing the protections offered to students and parents under the IDEA lest they sacrifice their funding.

What this means for speech-language services to special education students in California is that the IDEA basically says each eligible child must get whatever he/she needs in order to receive educational benefit, regardless of what type of services are required and regardless of the applicable eligibility categories. That’s the whole concept of individualizing a child’s education plan based on his/her unique educational needs.

There is nothing at the state-level that reduces this federal mandate, nor could there be unless California were to choose to go it alone to cover its special education costs and we all already know that California can’t pay its bills even with the federal funding it receives. It absolutely cannot afford to give up its federal special education funding.

We’re curious to know if there are any other state-level debacles involving misinterpretations of the law happening elsewhere. Readers are encouraged to post comments to this posting about such misinterpretations that may be occurring where they live.