re: shared attention and GrammarTrainer I had a funny experience a few days ago. Funny and wonderful.
I had gone to Andrew’s day program and logged him onto a new GrammarTrainer session. Then, when all seemed to be going well, I started reading my phone instead of looking at the laptop Andrew was using. (In theory, he’s supposed to use the program independently….)
And Andrew didn’t like it !
He poked me, protested (“Huh!“), and jabbed his pointer finger at the laptop screen. I thought he must have been having trouble with a question, but he wasn’t. He just wanted me to look at the same thing he was looking at.
I don’t think that’s ever happened before.
This may be a post only parents of autistic children and adults can really ‘get,’ but here goes.
One of the most painful aspects of autism for the parent (and no doubt for siblings and others close to the child) is the profound deficit in shared attention:
Joint [shared] attention occurs when two people share interest in an object or event and there is understanding between the two people that they are both interested in the same object or event.
Joint attention should emerge around 9 months of age and be very well-established by 18 months of age.
A 9-month old baby points. (Right? It’s been a while.)
A 9-month old baby points, and, when you point, s/he follows your finger to see what it is you’re pointing at. Parent and child look at the same thing at the same time, and they do so on purpose because people share.
But autistic babies don’t point. At least, neither of my autistic children pointed. Nor did they react when I pointed.
Turns out Andrew and I haven’t been following directions. (See: Syntax is not so easy.)
I wasn’t actually aware there were directions, but now that I know I still haven’t read and/or watched them.
This goes to one of the requirements of teaching apps–of any app–which is that people don’t RTFM.
If you need a beta tester to help with that, I’m your person.
This reminds me of a friend of mine whose husband was a composer with, she later suspected, the same learning issues their son had.
She once told me that his studio work was a marvel of intuitive button pushing.
Yesterday’s Sentence Weaver lesson featured the distinction between:
Which circles are biggest?
Where are the biggest circles?
The answers were things like:
“the circles in the middle [are biggest]”
“the biggest circles [are in the middle]”
He was not getting it at all.
Actually, I think he was getting one question & not the other, but I was too preoccupied trying to teach him what left-right-middle and top-bottom-middle meant to write down which question was which.
Anyway, I could almost hear him thinking:
These are all the same words!
What the hell?
Maybe I should be teaching him Latin instead.
I’m making headway on my New Year’s resolutions (7,000 steps a day for me, GrammarTrainer for Andrew, and possibly for Jimmy, too).
This morning I showed the folks at Andrew’s day program how he uses Katie’s program. They were amazed. Everyone is always amazed when they see Andrew using SentenceWeaver (must get videos loaded): thanks to SentenceWeaver, he is one of the few people on the planet who knows what a function word is.
This is a nonverbal person with severe autism. Knows he needs a function word to connect red to green when he’s saying an oval is red-and-green.
I’m still amazed myself, watching him.
I’m overdue for an “autism diaries” update–as a few recent developments with J have reminded me. At a time when the world as a whole seems so profoundly screwed up, it’s nice to reflect on how far this one little guy (now 6 foot 5) has come.
J was born profoundly deaf–so deaf that a brainstem test revealed no auditory awareness whatsoever. As this news took shape (literally, in the flattening lines of an EEG screen), we had no idea about cochlear implants. As far as we knew, J faced a future of profound silence.
That vision was soon moderated by cochlear implant pamphlets and phone calls to the parents of implanted toddlers. But just how well an implant would work for J in particular remained disconcertingly uncertain. There was, in the late 1990s, simply not enough data for firm predictions.
Some three months after J was implanted, I played a chord on the piano while he was facing away from me, standing on a chair, engaged in what we thought was a passing hobby: turning on and off a ceiling fan. I played the chord and he promptly turned around and looked right at me.
But why, the speech therapist wondered a few months later, did he remain so oblivious to speech? Enter diagnosis #2.
J as it turned out, was not just “profoundly” deaf, but “moderately” autistic. Not only that, but “moderate,” in the context of autism, was pretty bad. Unlike the “moderate” hearing loss we’d initially hoped for way back when, “moderate” autism (or so we were told) meant something much more debilitating: some language, minimal “splinter” skills, a lifetime of dependence.
Fast forward 19 years and, after several years of GrammarTrainer, an intense regimen of schooling (mostly in regular classes) and one-on-one tutoring by a variety of creative and talented lay people (and his parents), together with multiple-times-per-week outings all over the city and state, J is, according to two recent and thorough psychological exams, ….still moderately autistic.
But he’s also slowly making his way through college, majoring in computer science (and/or math). Though he lives at home, he independently gets himself to class every day, keeps track of assignments, exams, and schedule changes, meets with professors or TAs as needed, turns in his work, and participates in class and extra-curricular activities.
He’s even done some in-class presentations–and done OK on them. His grades aren’t perfect, and there’s nearly always a class or two that doesn’t work out the first time around and must be dropped at the last minute. But he’s managed to complete the English and Communications requirements, and has done well in two other courses you’d think would be huge struggles.
The first was linguistics. Assignments for this class included phonetic transcriptions of spoken English, and I doubted whether our profoundly deaf child could handle all the acoustic details. Would he hear the subtle air puff that accompanies the “p”-sound, but only when it occurs at the beginning of the word? Would he hear the difference between the “p” in “pat” vs. “spat”? I watched with amazement as J effortlessly completed these transcriptions, only then realizing how much of a miracle the cochlear implant really is:
The second course (still in progress) is psychology. You’d think a subject like this–with so much of its focus on emotions and social dynamics–would be particularly out of reach for a student with moderate autism. How accessible could concepts like oral fixation, or super-ego, or collective unconscious, or peer pressure, or social anxiety disorder, possible be?
As it turns out, there are two moderating factors. First, to the extent that psychology analyzes concepts that most of us have some intuitive understanding of, it makes those concepts more accessible to those who don’t. Second, a lot of psychology involves more accessible topics like neurology, cognition, and learning–some of which really resonate with J. Not surprisingly, he particularly enjoyed the chapter on cognitive illusions.
Still, it’s amazing to me to see (once I’ve laboriously worked through the textbook’s explanations with him) how quickly J grasps and generalizes concepts like operant and classical conditioning.
Just for the heck of it, I brought up this last topic in a recent exchange of text messages. I had headed out for the evening, having told J (when asked) that there would be no ceiling fans where I was going. A few hours later, he texted:
Where are you
I sent him this picture of the ceiling of my current location:
Are you blind?
Me, a bit confused:
Are you deaf?
I thought you said no fans
Were there fans in the picture? I’d somehow not noticed them. I looked up and scrutinized the ceiling and yes, sure enough, camouflaged among the black ceiling lights hanging from the black ceiling were some black ceiling fans.
Me (a bit sheepishly):
Well, they’re not moving
And they’re not going to move.
Ok but how many fans does the restaurant have?
And they will stay off.
I then decided it was my turn:
Now answer my question:
a. An unconditioned response
b. A conditioned response
c. An unconditioned stimulus.
d. A conditioned stimulus.
J (a short moment later):
Conditional on what?
I don’t remember.
It was years ago
It sure was.
Nature abhors a vacuum—particularly when it comes to to-do lists and worries. No sooner did I complete the big SentenceWeaver upgrade and deal with (at least for now) the various bugs that have sprung up during beta testing than I found myself worrying about a whole new issue: one that potentially undermines the entire program. This issue stems from a certain shadowy, world-wide organization that has the power to cause widespread disruption to websites.
No, it’s not Korean hackers. As far as hacking goes, I’m not particularly worried. My content is copyrighted; my code is encrypted; whole directories are blocked off from all IP addresses except mine. Then there are my web host’s gatekeeping algorithms, which are so risk-averse that they recently started blocking the IP address of my main beta-tester. A few of the thousands of words he’s typed in, as it turns out, appear on my host’s list of key words that could theoretically be used in attacking the website.
(This list includes “from” and “where”: words that appear regularly in the sentences that users input as part of their grammar training. The offending sentence, the one that got my user backlisted, was “The boy and the girl will wash the car three days from now.” Once I became aware of the issue, the solution was a simple string substitution before and after the php calls to the database.)
No, the shadowy world-wide organization to which I refer is the World Wide Web Consortium, aka the W3C. This is an organization of organizations, founded at the MIT lab for Computer Science “with support from” the European Commission and DARPA (the Defense Advanced Research Projects Agency). Its member organizations, which must be “reviewed and approved” by the W3C, range from businesses to universities to “governmental entities.”
The W3C’s mission, according to Wikipedia, is:
to foster compatibility and agreement among industry members in the adoption of new standards defined by the W3C. Incompatible versions of HTML are offered by different vendors, causing inconsistency in how web pages are displayed. The consortium tries to get all those vendors to implement a set of core principles and components which are chosen by the consortium.
In service of this goal, the W3C has adopted new standards for HTML5, the latest version of HTML that all browsers are eventually expected to use (think psychiatrists and the DSM V). One of these new standards involves “deprecating,” or no longer supporting, any HTML code that the W3C views as purely “presentational” in nature (think Asperger’s Syndrome).
One of these deprecated elements of HTML code is the lowly font tag—the tag used to specify aspects of font text like font type and color. In the words of the W3C’s website:
The element is a non-standard element.
HTML5 classifies it as a non-conforming feature.
HTML5-compliant websites are instead supposed to be handling color via Style Sheets.
If your website’s presentational elements are static, that’s fine. Indeed, in most websites, things like font type, font weight, and font color don’t change when you interact with the site. But one of the things that makes SentenceWeaver special—and is, in fact, an essential part of its Feedback Algorithm—is dynamically generated font color, as we see in this video below.
The prospect of my entire program, within the next few years, losing an essential part of its functionality, first kept me up at night—and then propelled me towards a workaround. Implementing it took me about a day and a half, and though the changes in code, in the end, probably summed to just a few extra lines, it was a kludgy pain in the neck.
One of the problems with shadowy, unrepresentative organizations inflicting rigid standards on the rest of us is their tendency to forget about unintended consequences. What we see here with the W3C, in particular, is a failure to imagine all the creative ways in which web tools can be used. Deprecate something, however lowly and insignificant it may seem to you, and suddenly algorithms you never thought to think about stop working, perhaps requiring many hours and kludges to rewrite.
The best defense of the W3C’s rigidity has to do with accessibility for people with special needs. The more rigid the standards for webpages, the easier it is to plug in accessibility tools like screen readers. But in my world, this is yet another example of accessibility at all costs—of ignoring the tradeoff between accessibility and remediation.
In my writings on disability in the classroom, I’ve worried that the emphasis on accessibility—along with the proliferation of assistive technology—has diminished the urgency of actual instruction. If students can communicate all urgent messages via picture buttons on tablets, why invest so many hours in teaching them to communicate with words?
The W3C standards put a different spin on this tradeoff: in prioritizing accessibility over website dynamics, they’ve undermined at least one program that caters to special populations as much in terms as instructional needs as in terms of accessibility.
Here’s a sample of SentenceWeaver’s Basic Pronoun Distinctions Module: