The "Homeopathic Fallacy" in Learning from Hypertext
The "Homeopathic Fallacy" in Learning from Hypertext 
McKendree, Reader and Hammon discuss three fallacies of hypertext: "Hypertext is like the brain," "Hypertext is like the mind" and "Structual similarity leads to educational effectiveness."
As the term "fallacy" suggests, the authors do not agree with these statements. They state that too little is known about the brain to allow such comparisons. And even if they would apply, there is no guarantee that this is the best media for learning.
The authors state that "hypertext is quite a good medium for searching, browsing and retrieving information," but avoid such general statement for learning. Instead, they provide simple recommendations: "Think active" and "use what people already know about text."
The former, think active, relies on the fact that learning is an active process; the latter, use what people know about text, on the fact that we already have learned to use linear text in it's various forms effectively. This knowledge should be used in hypertexts.
The article destroys the image that, with hypertext, we only need to copy the information directly into our brain, thus learning becoming a merely passive process. Learning is not simply a encoding process from linear text into hypertext.
"Hypertext breaks down the predictability found in linear text." That's true, we need new predictions for hypertext. Even for such simple things like the size of an article. For linear documents we have the number of pages, but what for hypertext?
Some parts are a little bit hard to read. Not because of the content, but in the way they are presented.
Some things are only explained superficially, like "hypertext is quite a good medium for ... retrieving information."
What are the three fallacies stated in this article?
Can we simply copy information from hypertext into our mind?