Optimal Encoding Method For Communication

By Lance McLemore, PRC-Saltillo Blogger

I’m not going to pretend to be unbiased here. I have strong opinions on this subject. However, I know in the field of AAC there are few absolutes and no one-size-fits-all solutions. What might be true and applicable to my situation is not necessarily true and applicable to another situation.

I was writing a blog post on another subject, but then I got on a Zoom call with John Halloran. He suggested that I write about spelling and word prediction versus semantic compaction as encoding methods. To be clear, semantic compaction refers to a method of representing language that is used in Unity and LAMP Words for Life. Basically, a symbol can have different meanings depending on how you use it.  I think I’m in a good position to write about it, since I’ve used both encoding methods.

For the first several years after I got my first AAC device, I used spelling and word prediction. To everyone, myself included, it seemed to be the best option as an encoding method. I was a literate grown man, so the choice seemed obvious. After I got my first AAC device, I didn’t really get that much use out if it. While there were multiple reasons for that, a major reason was that it was so inefficient. Let’s say for example that I wanted to type out the sentence, “I want to go eat now and get a hamburger.” It’s a relatively short trivial sentence. If I spelled it out on a keyboard it would require 41 keystrokes including spaces. I would consider 41 keystrokes to be a lot for such a simple utterance.

We want to be as fast as we can, so word prediction is an obvious potential rate enhancer. I type “I.” Then I type a “w.” A list of words pops up: will, where, what, were, when, or however many words will fit. I break my focus and look up at the list. I don’t see “want,” so I type an “a.” Finally I see “want” in the list, so I select it. I go through this process again and again as I type my message—constantly shifting my cognition from typing to searching for predicted words. If it’s not already obvious to you based on how I’ve described it, this process is quite exhausting. Such communication is not fluid and effortless. It’s a constant process of starting and stopping.

When I used word prediction, I estimate it would save me about 50% of my keystrokes. In the example sentence that’s 20.5 keystrokes. Most people think that’s impressive, but you need to look a little deeper. It saves me keystrokes, but you have to remember that I have to spend a significant amount of time searching the word prediction list after typing each letter. The result is that my keystrokes are reduced, but it might take me just as long or longer to type a message with word prediction as it would with spelling alone. On top of this, it’s very taxing on my brain, because I have to put so much thought and energy into the physical act of communicating. It leaves me with lower energy reserves to focus on the content of the message.

I use LAMP Words for Life which uses semantic compaction. I could write the above sentence in 16 keystrokes rather than 41. It’s faster than spelling alone, and faster than spelling and word prediction. Besides the greater speed, the other advantage which I consider just as important, is the reduced cognitive demand. Every word has one motor plan that never changes, so over time it’s become ingrained in my mind. As a result, I’m able to communicate almost as fast and easily as a speaking person.

As I’ve said before, there’s no one-size-fits-all solution for everyone. I don’t think Unity and LAMP Words for Life are the answer for every person, but why isn’t it considered as a viable option for many people? I think they would work very well for many people. I’m sure there are people like me who are literate, and people just assume that spelling is the best. Maybe there are other people who are emergent communicators or they have some cognitive impairment, and people just assume that a symbolic system is too confusing and overwhelming for them. Spelling is considered the ideal encoding method. In the field of AAC, there seems to be a hierarchy: spelling is the best followed by high tech symbolic systems and then PECS and so forth. If you spell, you’re considered the most intelligent and highest functioning. If you use a symbolic system, people see you as less intelligent and lower functioning.

I assure you that I can spell at least as well as the average person, but for communication I much prefer a symbolic system for the reasons I’ve already stated. When I was first introduced to LAMP Words for Life, I must admit that I was put off by it. I had the same prejudices that many people have. I assumed a symbolic system was inferior to spelling, and that people would see me as less intelligent and less capable if I used it. Maybe some people still think that, but I don’t really care anymore.

I wish that speech-language pathologists would be more open to introducing symbolic systems like LAMP WFL and Unity to their clients. I wish they would drop the assumption that by default a literate adult or child should use a spelling-based communication system. I wish they would understand that word prediction is not as great as they think; it’s only plausibly beneficial in theory. I wish they would see that a good communication system should place very little demands on cognition; using an AAC device should be as effortless as talking.