Saturday, February 18, 2017

Using Discourse Analysis to Assess Cognitive Decline

Figure from Gauthier et al. (2005).

Alzheimer's Disease (AD) and other dementias are progressive neurodegenerative conditions that unfold over time. Subtle symptoms such as forgetfulness and word finding problems may progress to mild cognitive impairment (MCI), and then escalate to full-blown dementia. Recent efforts to classify prodromal states have included automated analysis of spontaneous speech, which loses complexity as the disease progresses.

In one study, Frazier Fraser et al. (2015) applied machine learning methods to speech transcripts and audio files from the DementiaBank database. The participants were 167 patients with probable AD and 97 controls. The authors considered a total of 370 linguistic features, and found that a subset of 35 was able to classify patients vs. controls with 82% accuracy.1 While an advance over previous studies, this is not yet useful for diagnostic purposes. Another limitation was the relatively short length of the speech samples.2

Using factor analysis, the researchers found that four dimensions of speech3 were most indicative of dementia:
  • Semantic impairmentusing overly simple words
  • Acoustic impairment e.g., speaking more slowly
  • Syntactic impairment  using less complex grammar
  • Information impairment not clearly identifying the main aspects of a picture they were told to describe

Public figures who give repeated interviews leave a searchable record of spontaneous speech that can be analyzed for changes over time. Presidential press conferences provide another rich source of data for linguistic analysis.

Berisha et al. (2015) examined transcripts from the press conferences given by Ronald Reagan (1981-1989) and George H.W. Bush (1989-1993). We know that President Reagan received a formal diagnosis of Alzheimer's disease in 1994, five years after leaving office. And as far as we know, the elder Bush is still cognitively intact for his age (he's 92 now).

The quantified linguistic features included:
  • Number of unique words
  • Non-specific nouns – e.g., thing, something, anything
  • Filler words – well, so, basically, actually, literally, um, ah
  • Low-imageability, high frequency verbs – e.g., get, give, go, have, do

Reagan showed a significant decline in the number of unique words over the course of his presidency, but Bush did not.

Likewise, Reagan showed a significant increase in the use of non-specific nouns and fillers, but Bush did not.

There are several caveats here. Reagan was 69 when he was elected, while Bush was 64. Reagan was president for eight years and Bush for only four years; yet Bush held over twice as many press conferences as Reagan. Nonetheless, the results are consistent with a decline in cognitive function (which is not uncommon when aging from 69 to 77). Can we can classify Reagan as having MCI on the basis of these results? I don't think so. We'd really need comparable data from a population of demographically matched elderly participants.

President Donald Trump

After his Feb. 16 press conference, the public debate over whether President Trump is mentally unbalanced has intensified. Much of the current and past discussion has centered on the possibility of Narcissistic Personality Disorder (NPD), as speculated in The Atlantic and Vanity Fair and The Guardian. Sure, Trump has many of these qualities (that predate his actual grandiose status as POTUS):
  1. Grandiosity with expectations of superior treatment from others
  2. Fixated on fantasies of power, success, intelligence, attractiveness, etc.
  3. Self-perception of being unique, superior and associated with high-status people and institutions
  4. Needing constant admiration from others
  5. Sense of entitlement to special treatment and to obedience from others
  6. Exploitative of others to achieve personal gain
  7. Unwilling to empathize with others' feelings, wishes, or needs
  8. Intensely envious of others and the belief that others are equally envious of them
  9. Pompous and arrogant demeanor
And we can call him narcissistic in the generic sense of the word. But do we need to diagnose him with a quasi-psychiatric disorder, as in this NY Times letter signed by 35 mental health professionals?4
Mr. Trump’s speech and actions demonstrate an inability to tolerate views different from his own, leading to rage reactions. His words and behavior suggest a profound inability to empathize. Individuals with these traits distort reality to suit their psychological state, attacking facts and those who convey them (journalists, scientists).

In a powerful leader, these attacks are likely to increase, as his personal myth of greatness appears to be confirmed. We believe that the grave emotional instability indicated by Mr. Trump’s speech and actions makes him incapable of serving safely as president.

Dr. Allen Frances, chair of the DSM-IV task force, has forcefully argued that Trump does not meet criteria for NPD, because he is not distressed by his behavior:
Mr. Trump causes severe distress rather than experiencing it and has been richly rewarded, rather than punished, for his grandiosity, self-absorption and lack of empathy. It is a stigmatizing insult to the mentally ill (who are mostly well behaved and well meaning) to be lumped with Mr. Trump (who is neither).

Discourse Analysis

Here I'll suggest a different approach: can we quantify age-related neurological change using spontaneous speech?

“You know what uranium is, right?  It's this thing called nuclear weapons and other things.  Like, lots of things are done with uranium, including some bad things.  Nobody talks about that.  I didn't do anything for Russia.  I've done nothing for Russia.”

This is the most egregious example in the one hour, 17 minute train wreck. But there are other signs. He used the construction “very, very” 20 times. The word “thing” (and its variants) was uttered 102 times.

Am I going to diagnose him with anything? Of course not. That's unethical! But I will say that since Mr. Trump has been a public figure for nearly 40 years, we can objectively analyze his spontaneous speech and quantify any changes over time. I must emphasize that there is no magical scale to use for classification or comparison purposes (at least not yet). We don't know what's normal age-related decline and what's pathological.

I suggest that the best corpus of spontaneous speech data is the collection of Trump interviews/conversations with David Letterman. I believe they're unscripted, and there are many many of them on YouTube (I've linked to eight below). Letterman has aged too, so you might as well analyze his speech as well.


1 The authors performed...
...a 10-fold cross-validation procedure in which a unique 10% of the data (i.e., the ‘test set’) are used in each iteration for evaluation, and the remaining 90% (i.e., the ‘training set’) are used to select the most useful features (of the 370 available as described in “Features” above) and construct our models. The reported accuracy is an average across the 10 folds. In a given fold, data from any individual speaker can occur in the test set or the training set, but not both.
2 A show-stopping limitation is that the two groups were not matched for age or education. The mean age was 71.8 for AD vs. 65.2 for controls, and years of education 12.5 vs. 14.1 yrs.

3 See also Alzheimer’s Disease Markers Found in Speech Patterns (link via @aholdenj).

4 BTW, they're not supposed to diagnose non-patients, that's unethical.


Berisha V, Wang S, LaCross A, & Liss J (2015). Tracking discourse complexity preceding Alzheimer's disease diagnosis: a case study comparing the press conferences of Presidents Ronald Reagan and George Herbert Walker Bush. Journal of Alzheimer's Disease, 45 (3), 959-63 PMID: 25633673

Fraser, K., Meltzer, J., & Rudzicz, F. (2015). Linguistic Features Identify Alzheimer’s Disease in Narrative Speech. Journal of Alzheimer's Disease, 49 (2), 407-422 DOI: 10.3233/JAD-150520

Gauthier S, Reisberg B, Zaudig M, Petersen RC, Ritchie K, Broich K, Belleville S, Brodaty H, Bennett D, Chertkow H, Cummings JL. (2006). Mild cognitive impairment. The Lancet  367:1262-70.

Thomas, C., Keselj, V., Cercone, N., Rockwood, K., & Asp, E. (2005). Automatic detection and rating of dementia of Alzheimer type through lexical analysis of spontaneous speech. IEEE International Conference, 3, 1569-1574. doi: 10.1109/ICMA.2005.1626789

ADDENDUM (May 23 2017): I updated some of the dead video links. Also, two new posts have appeared today: Trump wasn’t always so linguistically challenged. What could explain the change? (by Sharon Begley at STAT) and Donald Trump: Cognitive decline or TDS? (by Mark Liberman at Language Log).

ADDENDUM #2 (Sept 30 2017): I updated the embedded video, but many of the new YouTube links have been taken down.

ADDENDUM #3 (Jan 6 2018): I again updated many of the new YouTube links.

Donald Trump on the David Letterman Show

11-10-1988 Letterman Donald Trump [corrected 9-30-17]

Donald Trump Interview on Letterman Show (1997) [corrected]

Donald Trump Interview on David Letterman Show (1998) [corrected]

Donald Trump on David Letterman Show (2008-08-08) [dead link]

Donald Trump talks business and banks on David Letterman Show (2009-02-18) [corrected]

Donald Trump on David Letterman Show (2010) [corrected]

Donald Trump on David Letterman 17 October, 2013 Full Interview

Donald Trump on David Letterman January 8th 2015 Full Interview [corrected]

You can find them all here here.

other Trump

The Trump Archive - over 900 televised speeches, interviews, debates, and other news broadcasts related to President Donald Trump. See post at Internet Archive.

Donald Trump 1980 Interview (Brokaw) [corrected]

Ronald Reagan videos

The President's News Conference - 1/29/81

The President's News Conference - 8/12/86

Iran/Contra Excerpt from 11/19/86

The President's News Conference - late Oct 1987


  1. You might also want to check out Gottschalk, L., Uliana, R., & Gilbert, R. "Presidential candidates and cognitive impairment: Measured behavior in campaign debates". Public Administration Review, 48 (1988) 613-619. The research uses the well- validate Gottschalk Gleser content scales to claim that Reagan was cognitively impaired as early as 1980.

    The popular press reaction is available here

  2. By the way, keep up your brilliant blog. It is the highlight of my day when I receive it.

  3. Bill - Thanks so much for the kind words. I also appreciate the links to the Gottschalk et al. study. I listened to brief clips of Reagan's later press conferences, and he sounded downright erudite and well-informed compared to Trump's Feb. 16 performance.

    1. I read an article suggesting Trump has symptoms or neurosyphillis. Any thoughts?

  4. I'm curious if people with ADHD also tend to use more non-specific and filler words too. Even those who are extremely curious (or have the ADHD+high curiosity combo).

    1. For what it's worth, I have ADHD and people have said I am very articulate. Of course, anecdotal evidence of one person does not a conclusion make.

  5. you might check the work of jonathan fine (deceased), professor of linguistics at Bar-Ilan University, Israel

    from his obituary

    Jonathan’s research focussed on text analysis and in particular on the analysis of transcripts of doctor/patient interactions as an entry to the interface of language and psychiatry, publishing extensively and showing a keen awareness of the fuzzy boundaries between psychiatry and neurology. He gradually developed a theoretical model for the analysis of psychatrist/patient interaction, culminating in his book, Language in Psychiatry: A Handbook for Clinical Practice (Fine, J. 2006. London: Equinox).

  6. Very interesting. And a great collection of useful links!

    The idea of cognitive decline would be plausible if not for the fact that his symptoms were always there. They are indicative of narcissistic psychopathy and a few learning problems, likely related to NP. He's slower and less careful with respect of image management, but that -- growing carelessness -- is the progression of narcissistic psychopathy (or malignant narcissism).

    Great blog, BTW.

  7. Elka - You seem rather certain. "Malignant narcissism" is not a formal psychiatric disorder. Have you done a quantitative analysis of the language in his past public appearances to rule out the possibility of cognitive decline?

    "Malignant narcissism is a hypothetical, experimental diagnostic category. Narcissistic personality disorder is found in the DSM, while malignant narcissism is not."

  8. The 82% accuracy ... was the 18% false positives or false negatives.

  9. "Frazier et al" -> "Fraser et al".

  10. Anonymous of April 1 - they didn't say, but I could have missed it.

    Anonymous of April 2 - Thanks for pointing out the typo, it's fixed now.

  11. Great piece, thanks. But I'm a little confused: I thought at the end there you were going to "analyze his spontaneous speech and quantify any changes over time." Did I miss something? Or were you just inviting others to do it?

    Not asking for diagnosis, of course, but the analysis would certainly be interesting.

  12. Mr. Holznagel - Thanks for your comment. I was indeed inviting others to do it, since computational linguistics is not my area of expertise.

    I suppose I could jump-start the enterprise by finding the best speech transcription software. Suggestions, anyone?

  13. This blog post was given the Y\The David Pakman Show by the blogger and was recently discussed there:

  14. Thanks, dustproduction. I became aware of the video on Sunday (April 2) after seeing a tweet from @HSchoenbacher. I hadn't heard of David Pakman before that. Overall, I appreciate the coverage and think he does a good job. But he overstates the case in his posts at Daily Kos and HuffPo. He begins,

    "A blogger has written a post claiming to show evidence that President Trump has a cognitive decline, possibly early stage dementia or Alzheimer’s. It may sound like the unhinged rants of a conspiracy theorist, but the blog post is actually quite sane."

    But I never actually claimed that Trump has dementia... Here's what did I say:

    "Am I going to diagnose him with anything? Of course not. That's unethical!"

  15. Hi Nurocrotic
    I think you could look into Mosaic Down's syndrome which gives simular results. I am not an expert it and it is a long shot but maybe worth talking to someone who is an expert.

    From a phycologis point of view he shows extreme lack of self awareness which is why he can lie and tell a Jewish reporter that he is the most antisemitic person he knows without batting a eyelid, most people could not do this. Down's kids desperately what to fit in and will pick up underlying issues in there peer group and magnify back to them, when changing peer groups they can do 180 flips to fit in. Lots of Down's kids end up in secure mental homes because they are easy to manipulate and set up by gangs to do crimes.

    DT was taught positive thinking, army trained and inherited 20 million and told he could run a business which may have help counter balance the typical defects of Down's syndrome.

    Typical mental

    Short attention span
    Poor judgmentImpulsive behavior
    Slow learning
    Delayed language and speech development
    Problems sleeping
    Loveing attention seeking

    Typical physical
    Flat nasal Bridge
    Small chin
    Flaps to eyes
    Upwards slanting eyes
    Short neck
    Pore mussel tone
    Small hands
    Flat hair
    Small mouth

    Donald Trump, Trump's Ali G and view from inside the white house?

    Positive Down's

    Early influence

    If I am right DT has fooled everyone and is quite impressive.
    PS this could be irresponsible to publish this to the press without more fact based research.


  16. Every news organisation that re-posts this lie is working for the deep state.

    This is how you will know the media that reports the truth from the media who work for the dark side, the ones who post this lie. They are traitors and all in Washinton that try to use this as a basis to get rid of the President, will be committing treason.

    Also be aware that there are witches....yes, actual witches casting spells on the President in an attempt to get him out of the Oval Office. Search the mainstream news and you will see.

    America, you are in the final time of grace. If Trump goes, followers of hell will no longer hide their dirty deeds. They will parade their evil in the light and you will be without excuse.

    R.I.P. Seth Rich

  17. I notice there are no pro-Trump comments in this section.

  18. Thanks for the article.

    Are people noticing significant differences in Trump's speech patterns between his 1988 and 2013 appearances on Letterman? (The 2015 interview doesn't appear to be available.)

    I watched each only once. Also, I'm a lay person but nothing stood out to me.

  19. Anonymous of May 16, 2017 1:32 PM -- It looks like quite a few of the videos have been taken down. Here's another version of the January 8th 2015 interview. He doesn't seem that bad, but I noticed a lot of repetitions. Here's his response to the Charlie Hebdo shooting the day before (which was never referred to as such):

    " look through thousands of years of history, you’ve always had people, and you’ve always had whether it’s crazy people or people with as you say a point of view. But this is.. uh, a this was a terrible thing. A lot of people killed and for no reason whatsoever. And it’s a tough, it’s a tough deal. I will say this. The uh magazine uh they’ve had a lot of problems over the years. And they were really you know we know about uh, we know about people that are very tough, and they were very tough on a lot of different people. But who would ever think who would ever think this would happen."

    Branching out from Letterman to other late night talk shows, Trump sounds more fluent (and funny) in friendly situations (e.g. with Jimmy Fallon) on topics he's comfortable with (Donald Trump on Board Games, His Health and Fast Food Habit) vs. ones that are more difficult (Donald Trump Clarifies His Relationship with Vladimir Putin).

    But overall, complete transcripts of these shows, as many as possible, are needed. Then a quantitative analysis of word usage would be necessary.

    1. It seems normal that stress would exacerbate the problematic speech and thought patterns in someone with early stages of Dementia. This could explain why "friendly" conversations would not appear to be abnormal.

  20. Except this study/page about NPD
    clearly states that someone with NPD does NOT show empathy, and the majority are NOT dismayed at all by what they do to others. It is considered by their victims to be one of the hallmarks of their behaviors. If they even notice, they don't care that they hurt you, it's your fault, obviously. So this ONE doctor saying he isn't NPD because he doesn't display empathy or "distress" about is full of BS. And if the man IS mentally ill, why should it belittle others who are also mentally ill? That's silly. I'm Bipolar, and saying he is NPD doesn't bother me, because it's either that or he's an outright psychopath. I grew up with both, I know what they look like and sound like and act like. He is smack on the spectrum.

    1. I couldn't agree more. The idea that people with NPD suffer from the disorder is flat wrong. That is why personality disorders are impossible to treat: the afflicted sees nothing wrong with their behavior and has no desire to change.

      I don't think anyone beyond partisan hacks could, at this point, believe that Donald Trump is mentally well.

  21. Anonymous of May 16, 2017 1:32 PM here. Thanks for your response. I agree quantitative analyses are needed. Also, it appears our speech is affected by multiple factors, stress being one.

    Today's article in Raw Story addresses your topic:

    The change in linguistic facility could be strategic; maybe Trump thinks his supporters like to hear him speak simply and with more passion than proper syntax.

    "He may be using it as a strategy to appeal to certain types of people," said Michaelis.

    But linguistic decline is also obvious in two interviews with David Letterman, in 1988 and 2013, presumably with much the same kind of audience. In the first, Trump threw around words such as aesthetically and precarious, and used long, complex sentences.

    In the second, he used simpler speech patterns, few polysyllabic words, and noticeably more fillers such as uh and I mean.

  22. Thanks for the link to that article, which originally appeared in STAT: Trump wasn’t always so linguistically challenged. What could explain the change?.

    Totally agree that stress (and sleep deprivation) can play a role too.

  23. Would any of his doctors, if they knew, have a responsibllity or even the power to deem him unfit to serve as Pres? Just curious on the role his person physian could have in diagnosising and treating any possible mental illness.

    1. My professional experience as a mental health professional for over 30 years is that MDs, with the possible exception of Psychiatrists, are focused almost entirely on physical symptoms. Additionally, when the personal physician has a long-term relationship with a patient, I believe they can become even less objective in diagnosing mental problems. I also believe that they frequently miss early symptoms of mental disorders, until those symptoms become more pronounced. It takes quite an extensive interview to diagnose Dementia, as well as substantial testing. That being said, I think there is much cause for concern about President Trump's mental condition.

  24. Although I admire Dr. Frances, I am FIRMLY on the other side of this thing. I have NEVER had a narcissistic patient who was distressed by their NPD. They cause a lot of distress in others. I would say that if Dr. Frances requires distress of the subject for his diagnostic criteria, look at the distress Trump feels when he is criticized. His lashing out can be seen as evidence of distress that is due to his NPD.

  25. Unknown - I can see your point here...

  26. I know a few men with Lewy Body Dementia, and I wonder if that's what Trump might suffer from, especially his needing to hold someone's hand while walking down a ramp (or stairs?), his paranoia, and possibly his insistence on the crowds at his inauguration could have been a visual hallucination. Also he seems to have insomnia (since he tweets so early in the morning), which is another common symptom.

    On the other hand, and even if he has dementia, he has had the grandiosity and lack of empathy all his life...

  27. The most critical question to answer regarding Trump's cognitive impairment using discourse analysis is the slope of the change and whether it has been relentlessly downward. Once the answers to these questions are known and the data are displayed graphically it becomes much less feasible to argue the validity of what is happening by contesting the specifics. One look at the graphic data for Ronald Reagan tells the tale.

    Additionally, once the rate of decline is established over a long baseline it becomes possible to make predictions regarding the future time course of Mr. Trump's cognitive deterioration and even to speculate about what the nature and severity of future problems is likely to be. At this point that is what should we really care about most.

    I would also point out that even very smnall changes in walking speed, 0.1 seconds more per year, were 47 percent more likely to predict subsequent cognitive decline (1). Because Mr. Trump has been a public person for so long it may be possible to do sophisticated gait analysis on the large archive of video footage that exists of him. Perhaps more to the point, Mr. Trump has exhibited such obvious signs of neurocognitive impairment that it should be easily possible to determine if he has one of the more common age-associated dementias. Early, noninvasive and highly accurate diagnosis of Mild Cognitive Impairment and Alzheimer's Disease are now a reality:

    What is not a reality is the presence of infrastructure that mandates the application of such objective tests (and others) to determine the fitness of a U.S. President to serve once there is evidence of cognitive decline. Perhaps in the future the need for such infrastructure will become more apparent and it will be implemented.

    1: Rosso AL, Verghese J, Metti AL, Boudreau RM, Aizenstein HJ, Kritchevsky S,
    Harris T, Yaffe K, Satterfield S, Studenski S, Rosano C. Slowing gait and risk
    for cognitive impairment: The hippocampus as a shared neural substrate.
    Neurology. 2017 Jul 25;89(4):336-342. doi: 10.1212/WNL.0000000000004153. Epub
    2017 Jun 28. PubMed PMID: 28659421.

  28. Hi guys,

    I was pretty taken by this idea until I saw the fits in the graphs above. I'm sorry but to me the fit in all the graphs isn't too convincing. The data points are all over the place. Using it as evidence to predict Reagan's decline - to me - feels like fitting the data to an outcome you already know to be true.

    Also, based on what you are doing, in my opinion, the x-axis of these plots should be time, not ordinal number of the speech. That could make a big difference, since the time frames are significantly different (4 years vs. 8 years).

    In short, based on what is presented here I'm not (yet) convinced of the premise that this form of speech analysis can predict neural degeneration with statistical significance.


  29. A person I know suffers from dementia caused by atypical multiple sclerosis. He also has difficulty with ramps.

  30. Informative post--I've collected some related material here:

  31. Trump's tax cut agenda is a serious business and must be understood properly.

  32. My mother was sharp; father was a regular columnist for the county paper. My sister and I recognize the dimentia 15 years prior our men's acknowledgement.

    Fortunately for them, there were no real medical advancements available at that time.

    I believe men may ignore, more than females, signs of stress. There are facial and visual cues, signals of distress you pick up. These folks can seemingly do pretty well for a while, they'll have moments of complete clarity. This is clarity is intermittent and the clarity could be terrifying.

    Both Mother and Trump prefer the night. Yes, I think the Dimentia is real.