I think they do fade in TEACCH, actually. Or at least they do in DD1's school.
For example, their resources are modified over time. When she first worked on 1:1 correspondence in maths, she had a laminated A3 card with the completed answer on the top half and velcroed numbers for her to place them in the right order -simply picture matching and identifying that the number looks the same. Once she could do that, it would change to reduce the prompt. Then it would have numbers missing. So all the time stretching what she could do. If she succeeded, they would look for consistency and move in. If she failed, they step back and try again.
In that sense it's no different from teaching NT children. They have a range of resources to make maths easier -number grids, number lines, multi-link.
Collecting the register-at first they went with her, modelled the request, returned to the classroom. Then they gradually reduced their input, then sent her on her own.
It's supposed to achieve reduced anxiety with a low arousal environment, clear instructions for tasks, visuals (created by someone else) for transitions, planning, life skills, limited language demands, individual desks to avoid stimulation from others and to avoid causing anxiety.
I wasn't slating ABA, but the way it is done in this country. Applies to TEACHH too.
What I don't like about TEACHH is that the strategies only reach a point. That point is when the person no longer needs a carer or person to help. It doesn't seem to go beyond that.
Prompts and reinforcers are not faded. They are seen to be doing a job and so they stay.
Also, I don't think being dependent on a set of visuals is independence. It might be as independent as a person can get in their lifetime, but for the many that can move on from them, well there isn't really a strategy to enable that.
Lastly, there is an implication in the UK, that TEACHH enables a child to be independent of a 1:1 or carer, or interventionalist, which is a person. I want my child to be DEPENDENT on people, as a NT child would be, to learn, garner information, seek directions, interact, develop a relationship, - as much as they are able and without ever giving up.
I can't get too worked up about it all. I think people who practice TEACCH tend to have misconceptions about ABA (robotic learning, enforced learning with no fun...blah blah) and people who practice ABA have misconceptions about TEACCH (no direction, workstations, pretty visuals, no stretching...blah blah).
The truth is, either can be done well or badly. The proof of the pudding is in children progressing and meeting their potential.
I think it all depends on the person using it. Teachers who use teaching methods like prompts and reinforcement cannot do so effectively without assessing individual student interests and learning styles. So unless the person using it has taken the time to observe the Dc or use any information available to them about the Dc in their care I don't think it would be as effective.
It's not a BAD system. It was amazing when it first came in as it gave PeoPle with ASD a voice and a level of independence from carers.
However, it hasn't been brought up to date nor developed in response to alternative research-based interventions.
But that isn't why I woukd avoid it. My no. 1 reason is that I have never come across a school or unit that actually uses TEACHH despite pretending to. What I see is workstations and visuals and child-occupying tasks, which is paraded as TEACHH but actually just a small, cheap, part of it.