Graph analysis of dream reports
January 19, 2014 9:14 AM   Subscribe

Elsa: [reading Jacob's palm] "See. According to this, you're already dead."
posted by clavdivs at 9:39 AM on January 19, 2014 [3 favorites]

The confounder evident to me is the facility for recapitulating poorly remembered events. I wish they had asked all three groups to describe a poorly remembered book, film, or play to compare to how they describe their dreams. What's implicated here is both the character of the subjects' dreams (what the paper makes an implicit conclusion about when it validates dream analysis) and how they describe those dreams.

The researchers used a description of events experienced before sleep as the "waking" narrative against which the dream narrative is compared, but the events experienced before sleep are of a very different character than dreams.

This is of interest to me because I am atypical in my sleep patterns and dreaming, with a large portion of my dreaming in the form of coherent, non-personal narratives such as in stories, usually similar to the kinds of novels and films I read — that is, frequently fantastical or crime stories or other genre narratives. In these stories, I'm usually not myself. There is nevertheless some degree of incoherency and discontinuity, those this varies. Also, I have a huge repertoire of recurrent themes, particular storylines, and especially fictional places. I have a few dream versions of real places I've been that, through years and years of dream recurrence, have swamped some of my real memories of those places and sort of exist in a ghost form alongside them in my waking recollection. Which is a little confusing.

All this to say that I suspect that my dream graphs wouldn't look quite like any of those three examples — they'd be more dense and less linear than the schizophrenic graphs, but more linear and less dense than the normal and bipolar graphs (and probably somewhere between a normal dream graph and a normal recapitulated movie graph).

But, also, I think that any clinician would almost certainly be able to diagnose schizophrenics and manic-depressive (in the manic stage) and normal people via a recounting of dreams to a similar reliability.
posted by Ivan Fyodorovich at 10:02 AM on January 19, 2014 [5 favorites]

Hmmm. How do they control for people that (like me) have kept a dream journal and lucid dream regularly, thereby increasing the detail, amount and linearity of dream recall?
posted by 3FLryan at 10:47 AM on January 19, 2014

I object violently to their conclusion, the same results could have been obtained by asking people what they did yesterday. Noticing these speech patterns is bog standard differential diagnosis.
posted by hobo gitano de queretaro at 10:59 AM on January 19, 2014 [1 favorite]

"...the same results could have been obtained by asking people what they did yesterday."

That's what they did, basically. The point is that comparing those accounts of "what they did yesterday" (what they did before they went to sleep, actually) against their accounts of the dream they had, shows differences between waking and dreaming accounts and between the categories of subjects.

They did account for basic speech patterns, both by asking to recount the events before the dream, and also by trying to remove simple verbosity as a variable, and also by using the method on these accounts as translated into other languages.

In other words, they did make a good effort to not look at what you call "speech patterns" but to look at the narrative content — in its variability, linearity, and connectivity — by abstracting that content in the form of particular words and their connections to other words, thus the graphs.

My criticism that's similar to yours is that these differences in content in the spoken accounts would almost certainly be an obvious and sufficient diagnostic marker for a clinician, anyway.

My other criticism is that differences in styles of recapitulation of a complicated and poorly-remembered narrative (which dreams usually are, and which the events before bedtime usually are not) could account for this, and not the content of the dreams themselves, as the authors suppose.
posted by Ivan Fyodorovich at 11:13 AM on January 19, 2014 [3 favorites]

Ivan you are a smart dude and I honestly salute you for being able to articulate ideas like that.
posted by Teakettle at 12:46 PM on January 19, 2014

« Older To be a good astronaut, you need to be prepared...   |   When Artworks Crash Newer »

This thread has been archived and is closed to new comments