Losing track of words

November 3, 2011 10:51 AM

Can vocabulary analysis detect the onset of Alzheimers Disease in writers? In 2004, a team at UCL demonstrated that Iris Murdoch's last novel had simpler sentence structure and a smaller vocabulary than her earlier books. Now a team at the University of Toronto has corroborated that research, and suggests that Agatha Christie too suffered from the disease at the end of her career.
posted by Sonny Jim (37 comments total)

Also covered by Radiolab.
posted by bjrn at 10:57 AM on November 3, 2011

Terry Pratchett would be a good person to look at...
posted by Chuffy at 10:58 AM on November 3, 2011

Huh, I thought I'd read this here before but it must have been a different via. That paper was presented in 2009, so it's not really "now" that the team has corroborated that research.
posted by DU at 10:59 AM on November 3, 2011

This was part of a fascinating story on Radiolab some time back.

Vanishing Words

More detail at NPR
posted by maudlin at 11:00 AM on November 3, 2011

(Same story, different broadcast date as linked by bjrn above.)
posted by maudlin at 11:00 AM on November 3, 2011

And ever since reading about this, I've wondered if you can do the same with the spoken word. It's going to depend on who you are talking to, of course.

Oh wait, except you don't need voice on the internet. A browser(/im/irc/usenet/email) plugin could monitor it for you over time, averaging the values to reduce the "I talk like a child to idiots on usenet" effect.
posted by DU at 11:04 AM on November 3, 2011 [2 favorites]

simpler sentence structure and a smaller vocabulary

So...Dan Brown, Mitch Albom, Stephen King and Tom Clancy all have Alzheimers?
posted by R. Schlock at 11:04 AM on November 3, 2011 [2 favorites]

Update to Siri to detect diminished complex ways in which to query where one can dispose of a body in 5 ... 4 ... 3 ...
posted by PapaLobo at 11:06 AM on November 3, 2011

DU: sure, the article's based on research that's already been presented at conferences, but this would be its official, peer-reviewed write-up. Scholarly publishing and its lagging ways, etc. Bjrn and maudlin: thanks for linking that Radiolab programme. I'll have a dip into it now.
posted by Sonny Jim at 11:09 AM on November 3, 2011

Can vocabulary analysis detect the onset of Alzheimers Disease in writers?

me dont think so
posted by sexyrobot at 11:15 AM on November 3, 2011 [3 favorites]

I always sort of suspected this with Frank Herbert, given how he appeared to lose understanding and control of his own setting and characters in the last couple of Dune books. Of course, his son and Kevin J. Anderson don't have this excuse...
posted by Strange Interlude at 11:16 AM on November 3, 2011 [1 favorite]

Terry Pratchett would be a good person to look at...

I don't know. Pratchett has Benson's syndrome which doesn't seem to affect language.
posted by Pendragon at 11:18 AM on November 3, 2011 [1 favorite]

gah, I should read better. From the wikipedia article I linked to:
By contrast, individuals with posterior cortical atrophy tend to have well preserved memory and language but instead show a progressive, dramatic and relatively selective decline in vision and/or literacy skills such as spelling, writing and arithmetic.
posted by Pendragon at 11:21 AM on November 3, 2011

Pratchett dictates his books now. I wonder how that splits the cognitive line between speaking and writing?
posted by CaseyB at 11:30 AM on November 3, 2011

But how specifically Alzheimer's and not another type of dementia?
posted by inturnaround at 11:52 AM on November 3, 2011

A really neat concept. The idea that doctors' intuition can be systematized to diagnose diseases early is incredibly helpful.

This puts me in mind of developments for diagnosis for Parkinson's Disease: looking at how someone walks. The gait of someone with Parkinson's seems to be similarly characteristic:

PD patients walked slower with shorter stride-length, comparable cadence, and longer double support times. Kinematics showed a reduction of the range of motion in the hip, knee, and ankle joints. Maximum hip extension and the ankle plantar flexion were significantly reduced. Kinetic gait parameters showed reduced push-off ankle power and lift-off hip power generation. Strong correlations between these important body advancement mechanisms and the walking velocity were observed.

posted by bonehead at 11:57 AM on November 3, 2011

See also: The Nun Study of Aging and Alzheimer's Disease

http://en.wikipedia.org/wiki/Nun_Study

It seems that one's baseline writing style is a predictor, not just changes.
...the approximate mean age of the nuns at the time of writing was merely 22 years. Roughly 80% of nuns whose writing was measured as lacking in linguistic density went on to develop Alzheimer's disease in old age; meanwhile, of those whose writing was not lacking, only 10% later developed the disease.
posted by SLC Mom at 12:01 PM on November 3, 2011 [2 favorites]

If you applied this theory to Cormac McCarthy the program would probably determine he had a stroke some time in between publishing Suttree and Blood Meridian.
posted by nathancaswell at 12:19 PM on November 3, 2011 [1 favorite]

last novel had simpler sentence structure and a smaller vocabulary than her earlier books.

I would consider that to be better writing, the result of years of experience.

Hemingway had it right.
posted by Tell Me No Lies at 1:18 PM on November 3, 2011 [1 favorite]

Sonny Jim: “Now a team at the University of Toronto has corroborated that research, and suggests that Agatha Christie too suffered from the disease at the end of her career.”

Not just at the end of her career, either.

“Hrm, what shall I call this latest awful detective story? ‘Ten Little... Ten Little...’ HOLD ON, I'VE GOT IT!”

That said, while I can support any theory that ascribes dementia to Agatha Christie, this one doesn't make any sense. Writers sometimes use simple sentence structure, and sometimes more complex. Writers who use simpler sentences than other writers are not necessarily Alzheimer's sufferers. The same goes for vocabularies. The idea that these things are correlated seems laughable to me when there are clearly reasons why a writer might want to simplify sentence structure and vocabulary on purpose.
posted by koeselitz at 1:35 PM on November 3, 2011 [1 favorite]

SLC Mom: “It seems that one's baseline writing style is a predictor, not just changes. (link)”

No – that's not at all the case.

The Nun Study has shown that Alzheimer's disease can be a predictor of one's writing style. It has not shown that writing style can be a predictor of Alzheimer's. That's an essential and fundamental difference that this research on Iris Murdoch and Agatha Christie utterly fails to take account of.
posted by koeselitz at 1:39 PM on November 3, 2011 [1 favorite]

I do not want to know anything more about predicting Alzheimer's disease until we know how to treat it.
posted by Maias at 1:50 PM on November 3, 2011 [3 favorites]

This should be known as a Reverse Henry James.
posted by villanelles at dawn at 1:50 PM on November 3, 2011 [1 favorite]

I would consider that to be better writing, the result of years of experience.

Hemingway had it right.

What's right for Hemingway isn't necessarily right for Iris Murdoch (or Woolf or Proust or Wallace or). She may have used more words, but that doesn't mean there was anything to cut.
posted by Adventurer at 2:10 PM on November 3, 2011

koeselitz: ... while I can support any theory that ascribes dementia to Agatha Christie, this one doesn't make any sense. Writers sometimes use simple sentence structure, and sometimes more complex. Writers who use simpler sentences than other writers are not necessarily Alzheimer's sufferers. The same goes for vocabularies. The idea that these things are correlated seems laughable to me when there are clearly reasons why a writer might want to simplify sentence structure and vocabulary on purpose.

OK, healthy scepticism is perhaps warranted. But Lancashire et al. are careful to corroborate their findings with other evidence. Christie's biographer cites clear examples that the author was deteriorating mentally in her last years, and had trouble completing manuscripts and composing plots. Alzheimers and other forms of dementia have clear, measurable impacts upon a sufferer's linguistic abilities, memory, and cognitive ability. Isn't it more parsimonious to propose that the measured decline in Christie's vocabulary in her late books, combined with their increased numbers of verbal repetitions and indefinite words, correlate with these known biographical facts, rather than being purely coincidental aesthetic choices? And they establish a similar pattern in Murdoch's writing (the fact of whose Alzheimers was confirmed during her autopsy), which establishes a further correlation.

So, sure: limited vocabulary size and simple sentence structure can be conscious choices. But when placed in a wider, biographical context of demonstrated mental decline, they can also indicate something else. I don't see why it's "laughable" to make that claim at all.
posted by Sonny Jim at 2:11 PM on November 3, 2011

Huh, I thought I'd read this here before but it must have been a different via. That paper was presented in 2009, so it's not really "now" that the team has corroborated that research.

DU,
Deja vu here too.

I'm sure I remember from the film Iris (2001), based on the memoir by her partner (& writer/academic John Bayley) Murdoch herself was shown as vividly/horribly aware that her writing powers were going. I get the point that this has now been supported by research, but it was a jolting moment in the film - a writer realizing another novel was not going to be possible.
posted by Jody Tresidder at 2:24 PM on November 3, 2011 [1 favorite]

Tell Me No Lies: "last novel had simpler sentence structure and a smaller vocabulary than her earlier books.

I would consider that to be better writing, the result of years of experience.

Hemingway had it right."

There's a noticeable difference between terse, tight sentence structure and simply not having the skill or vocabulary to express yourself as effectively any more, though.

I first noticed Terry Pratchett's tone changing in Monstrous Regiment, which I felt was far darker than his earlier stuff, probably as a side effect of his own outlook having grown more bleak. But after that came some truly glorious books, like Going Postal and Making Money, when he was at the top of his form. It took Nation, ostensibly a young adult book not set in the Discworld, for him to reach that dark place again. The point is, the skill was evident throughout.

It wasn't until Unseen Academicals that I felt Pratchett's writing changed. He always wrote long, complex sentences full of parenthetical expressions, referenced by not one but multiple footnotes--but now the parenthetical expressions meandered away from the topic at hand and didn't always find their way back again. The book felt...messier. Reading it felt like when you tag along with someone who swears he knows a shortcut to get you where you want to go, but ends up getting you lost instead.

And then Pratchett started dictating his work, and perhaps as a result of that, or maybe because of the progression of his illness, his sentences grew shorter and less complex.

Which would be fine, if he were at the top of his form and just "omitting needless words."

But at the point in his books where he used to deliberately tighten up the structure--at the denouement, when his characters rise to the occasion and words become actions--his work lacks the powerful punch of purpose it once had.

You'll still see those flashes of brilliance, of course, but whereas Pratchett's classic works are chock-full of perfect prose, comic genius that piles up, layer on top of layer, until you laugh out loud, more often than not now you feel like he's resorting to repeating old gags, like a favorite TV series that's run out of fresh writing ideas, says, "Fuck it," and takes the lazy route with a "best of" retrospective instead.

I'm (finally!) reading Snuff right now, and the old "wife telling the husband he doesn't really like eating something that's bad for his health" running gag is driving me batty, it is being hammered so hard.

And so far, there is no sign of Death, either. Who only showed up very briefly in Unseen Academicals as well. Understandable, for a man who has been forced to look his own mortality in the face, but I miss Death.
posted by misha at 2:29 PM on November 3, 2011 [1 favorite]

I wish they had done this with more than just 'Jackson's Dilemna'. Some of her fans claim there's a noticeable slip in her last three books.
posted by BigSky at 3:38 PM on November 3, 2011

I wonder if this can be applied to fan fic.
posted by ZeusHumms at 3:40 PM on November 3, 2011

uhh i think fan fic itself is a... sign of... alzheimers? shit. come on work with me here
posted by This, of course, alludes to you at 4:31 PM on November 3, 2011

I think alzheimers would be one of the least disturbing revelations provided by a thorough analysis of fan fic.
posted by Dark Messiah at 4:40 PM on November 3, 2011

yeah thats what i was going for, thanks dark messiah
posted by This, of course, alludes to you at 4:44 PM on November 3, 2011

You know, you could imagine this as a future feature of a personalized 'Siri'. As 'my' Siri gets to know me over the years, and learns what my typical speech patterns and vocabulary are like, it would not be too far-fetched to imagine that it could have the intelligence to recognize such changes in my conversation. It would be a 'mental health' version of those future toilets that are testing my urine every morning for physical health problems.
posted by woodblock100 at 9:11 PM on November 3, 2011

Writers who use simpler sentences than other writers are not necessarily Alzheimer's sufferers.

I thought the comparison was within a writer's own work, not between different authors. That's what I got from the Iris Murdoch and Agatha Christie situations, at any rate.
posted by harriet vane at 9:24 PM on November 3, 2011

I would consider that to be better writing, the result of years of experience.

Lightning v. Lightning bug. A loss of vocabulary doesn't necessarily mean a transition from the ornate to the plain; could be one from the plain to the simplified. Hemingway may have been short and crisp and Anglo-saxon, but a lot of his power comes from exactitude, from juxtaposition and allusion and connotation. Simple words, but exactly right, and powerful because so much is left unspoken. Simplicity is merely that without exactitude.
posted by Diablevert at 5:25 AM on November 4, 2011

Henry James got more complex - but that could be due to the fact that his later books were dictated to scribes, the earlier ground out by hand.
posted by IndigoJones at 3:57 PM on November 4, 2011

misha, I'm glad you said that, because I felt bad just thinking about it. Snuff had plenty of brilliance, but I don't think it was put together as well as other books. Pratchett seemed to be saying goodbye to his characters, and I thought the ending was too, well -- anyhow, I noticed the same thing while reading I Shall Wear Midnight and Unseen Academicals. Still, I was embarrassed that this made me think of his Alzheimer's. I thought to myself, "Why don't you just decide that he's moving in a different direction than what you would prefer; why do you have to pathologize it?"
posted by Countess Elena at 7:16 PM on November 4, 2011 [1 favorite]