The author of the OP article, Tanya Luhrmann, is well known and has done extensive research in cultivated halluncinations. These are not necessarily religious in character. She helped Stanford students experience Leland Stanford, Jr as a a presence in their lives. Although developing the skill to expereince secular but meaningful hallucination was suggested to her by a colleague, the idea is not original with her. She is, however, innovative in studying the phenomenon within a controlled experimental investigative design.
http://uncertaintist.wordpress.com/...t-thinking-about-thought-forms-from-stanford/
In other words, Professor Luhrmann's larger point is that hallucinations are not only common, but having hallucinations is an easily learned skill. A few years ago, as a personal experiment and before I knew anything about her work, I taught myself to have a specific (and I think "safe") closed-eyes visual hallucination. It is the "afterimage" you would see if you looked at a black letter "X" on a brightly lit white field for a minute or so. I mastered the skill inside of an hour. I still do it almost every day.
That I learned it and have maintained the skill is "proof of comcept." It is a skill, however strange that might sound, and so its exercise is not necessarily a clinical symptom. Other people want more ambitious hallucinations than I do, in other sensory modalities, The literature richly attests that they can learn to have those, too, whether self-taught or taught by someone else.
I wasn't interested in anything more ambitious. Just about any skill is subject to dissociative performance, which concerned me. That is, mastered skills will often operate without conscious direction. A familiar example is that so many people can drive "without thinking about it," and yet drive successfully.
Just the other day, for example, I was thinking about a document handling problem, and decided I would place a yellow-fluorescent "X" marking on certain papers. I didn't do anything about it, I just thought about the problem.
I then happened to close my eyes, and there was my now-familiar after-image hallucination, vividly displayed, attesting to the operation of the mechanism which I had developed. I had elicited the image, and its appearance was meaningfully related to my conscious activity, but I hadn't "invoked" it nor "asked for it." It kicked in "on its own."
If I had been interested in something more ambitious, it is reasonably clear that I could have had an autonomously appearing God, or an autonomously appearing Leland Stanford, Jr. Well, OK, if I had been interested in that, then it would be Keira Knightley. But the principle is well established. Halucinations are not necessarily of pathological origin.
If I may add some remarks on other comments,
My point in the OP quote is that the article's author appears to be making a logical fallacy: hearing god's voice isn't crazy, because it's common.
That is not the article author's position, and what Professor Luhrmann says, she can back up with extensive field work and experimental investiagtion, much of it prominently published, peer-reviewed and all of it thoroughly mainstream anthropology and psychology. Perhaps the article did not fully develop her argument, but this can hardly be surprising in a popular and non-scholarly outlet.
The anthropology of mental health practice is, by the way, another of Professor Luhrmann's research areas, in addition to "religious" expression. Therefore, I am confident that she, more than most people, could do without an introductory level exposition of "logical fallacies" as they apply to the problem of defining mental illness, with all due respect. Your later ackowledgment of Professor Luhrmann's cogency is noted with approval.
The article goes on to talk about situations I wouldn't characterize as "hearing voices".
However, what the author means by "hearing voices" combines at least two separable phenomena: a halucinatory experience and the dissociative quality of the experience. The combination is what interests her. So, it is not just that the subject "sees" or "hears" Leland Stanford, Jr, but also "interacts" with him, not as a ventriloquist interacts with a dummy, but rather as one sentient being interacts with a different sentient being.
Convincing myself that he was actually talking to me is the part it gets "crazy", though I don't like that word here, as it always conflates "mental disfunction" with "stupid or bad idea".
That would be a third possible component to this kind of experience, the failure to connect the skill-building phase (the effort to learn to encounter Mark Twain, in your example) with the subsequent productive exercise of the skill.
Luhrmann's research is ethical (and human-subjects regulated!). A religious entrepreneur would not neessarily disclose to "subjects" the potential of learning hallucinatory skills from engaging in well-crafted "devotional" activities. Basically, that would parallel a malefactor secretly putting a hallcinogenic drug in somebody's drink.
In such a case, the victim is not reasonably suspected of mental illness because they fail to figure out what was done to them, and especially not for failing to figure it out while the experence is unfolding.