Contribute Try STAT+ Today

It was the kind of utterance that makes professional transcribers question their career choice:

“ … there is no collusion between certainly myself and my campaign, but I can always speak for myself — and the Russians, zero.”

When President Trump offered that response to a question at a press conference last week, it was the latest example of his tortured syntax, mid-thought changes of subject, and apparent trouble formulating complete sentences, let alone a coherent paragraph, in unscripted speech.


President Trump denied his campaign colluded with Russia while speaking at a press conference in May 2017. Via YouTube

He was not always so linguistically challenged.

STAT reviewed decades of Trump’s on-air interviews and compared them to Q&A sessions since his inauguration. The differences are striking and unmistakable.


Research has shown that changes in speaking style can result from cognitive decline. STAT therefore asked experts in neurolinguistics and cognitive assessment, as well as psychologists and psychiatrists, to compare Trump’s speech from decades ago to that in 2017; they all agreed there had been a deterioration, and some said it could reflect changes in the health of Trump’s brain.

In interviews Trump gave in the 1980s and 1990s (with Tom Brokaw, David Letterman, Oprah Winfrey, Charlie Rose, and others), he spoke articulately, used sophisticated vocabulary, inserted dependent clauses into his sentences without losing his train of thought, and strung together sentences into a polished paragraph, which — and this is no mean feat — would have scanned just fine in print. This was so even when reporters asked tough questions about, for instance, his divorce, his brush with bankruptcy, and why he doesn’t build housing for working-class Americans.

In an interview from 1987, Donald Trump talks about poverty and homelessness in the US. Via YouTube

Trump fluently peppered his answers with words and phrases such as “subsided,” “inclination,” “discredited,” “sparring session,” and “a certain innate intelligence.” He tossed off well-turned sentences such as, “It could have been a contentious route,” and, “These are the only casinos in the United States that are so rated.” He even offered thoughtful, articulate aphorisms: “If you get into what’s missing, you don’t appreciate what you have,” and, “Adversity is a very funny thing.”

Now, Trump’s vocabulary is simpler. He repeats himself over and over, and lurches from one subject to an unrelated one, as in this answer during an interview with the Associated Press last month:

“People want the border wall. My base definitely wants the border wall, my base really wants it — you’ve been to many of the rallies. OK, the thing they want more than anything is the wall. My base, which is a big base; I think my base is 45 percent. You know, it’s funny. The Democrats, they have a big advantage in the Electoral College. Big, big, big advantage. … The Electoral College is very difficult for a Republican to win, and I will tell you, the people want to see it. They want to see the wall.”

For decades, studies have found that deterioration in the fluency, complexity, and vocabulary level of spontaneous speech can indicate slipping brain function due to normal aging or neurodegenerative disease. STAT and the experts therefore considered only unscripted utterances, not planned speeches and statements, since only the former tap the neural networks that offer a window into brain function.

The experts noted clear changes from Trump’s unscripted answers 30 years ago to those in 2017, in some cases stark enough to raise questions about his brain health. They noted, however, that the same sort of linguistic decline can also reflect stress, frustration, anger, or just plain fatigue.

Ben Michaelis, a psychologist in New York City, performed cognitive assessments at the behest of the New York Supreme Court and criminal courts and taught the technique at a hospital and university. “There are clearly some changes in Trump as a speaker” since the 1980s, said Michaelis, who does not support Trump, including a “clear reduction in linguistic sophistication over time,” with “simpler word choices and sentence structure. … In fairness to Trump, he’s 70, so some decline in his cognitive functioning over time would be expected.”

Some sentences, or partial sentences, would, if written, make a second-grade teacher despair. “We’ll do some questions, unless you have enough questions,” Trump told a February press conference. And last week, he told NBC’s Lester Holt, “When I did this now I said, I probably, maybe will confuse people, maybe I’ll expand that, you know, lengthen the time because it should be over with, in my opinion, should have been over with a long time ago.”

In an interview conducted earlier this month, President Trump explains the timing of James Comey's firing. Via YouTube

Other sentences are missing words. Again, from the AP: “If they don’t treat fairly, I am terminating NAFTA,” and, “I don’t support or unsupport” — leaving out a “me” in the first and an “it” (or more specific noun) in the second. Other sentences simply don’t track: “From the time I took office til now, you know, it’s a very exact thing. It’s not like generalities.”

There are numerous contrasting examples from decades ago, including this — with sophisticated grammar and syntax, and a coherent paragraph-length chain of thought — from a 1992 Charlie Rose interview: “Ross Perot, he made some monumental mistakes. Had he not dropped out of the election, had he not made the gaffes about the watch dogs and the guard dogs, if he didn’t have three or four bad days — and they were real bad days — he could have conceivably won this crazy election.”

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.”

Donald Trump shares his take on Ross Perot's 1992 presidential campaign. Via YouTube

The reason linguistic and cognitive decline often go hand in hand, studies show, is that fluency reflects the performance of the brain’s prefrontal cortex, the seat of higher-order cognitive functions such as working memory, judgment, understanding, and planning, as well as the temporal lobe, which searches for and retrieves the right words from memory. Neurologists therefore use tests of verbal fluency, and especially how it has changed over time, to assess cognitive status.

Those tests ask, for instance, how many words beginning with W a patient can list, and how many breeds of dogs he can name, rather than have patients speak spontaneously. The latter “is too hard to score,” said neuropsychologist Sterling Johnson, of the University of Wisconsin, who studies brain function in Alzheimer’s disease. “But everyday speech is definitely a way of measuring cognitive decline. If people are noticing [a change in Trump’s language agility], that’s meaningful.”

Although neither Johnson nor other experts STAT consulted said the apparent loss of linguistic fluency was unambiguous evidence of mental decline, most thought something was going on.

John Montgomery, a psychologist in New York City and adjunct professor at New York University, said “it’s hard to say definitively without rigorous testing” of Trump’s speaking patterns, “but I think it’s pretty safe to say that Trump has had significant cognitive decline over the years.”

No one observing Trump from afar, though, can tell whether that’s “an indication of dementia, of normal cognitive decline that many people experience as they age, or whether it’s due to other factors” such as stress and emotional upheaval, said Montgomery, who is not a Trump supporter.

Even a Trump supporter saw and heard striking differences between interviews from the 1980s and 1990s and those of 2017, however. “I can see what people are responding to,” said Dr. Robert Pyles, a psychiatrist in suburban Boston. He heard “a difference in tone and pace. … What I did not detect was any gaps in mentation or meaning. I don’t see any clear evidence of neurological or cognitive dysfunction.”

Johnson cautioned that language can deteriorate for other reasons. “His language difficulties could be due to the immense pressure he’s under, or to annoyance that things aren’t going right and that there are all these scandals,” he said. “It could also be due to a neurodegenerative disease or the normal cognitive decline that comes with aging.” Trump will be 71 next month.

Northwestern University psychology professor Dan McAdams, a critic of Trump who has inferred his psychological makeup from his public behavior, said any cognitive decline in the president might reflect normal aging and not dementia. “Research shows that virtually nobody is as sharp at age 70 as they were at age 40,” he said. “A wide range of cognitive functions, including verbal fluency, begin to decline long before we hit retirement age. So, no surprise here.”

Researchers have used neurolinguistics analysis of past presidents to detect, retrospectively, early Alzheimer’s disease. In a famous 2015 study, scientists at Arizona State University evaluated how Presidents Ronald Reagan and George H.W. Bush spoke at their news conferences. Reagan’s speech was riddled with indefinite nouns (something, anything), “low imageability” verbs (have, go, get), incomplete sentences, limited vocabulary, simple grammar, and fillers (well, basically, um, ah, so) — all characteristic of cognitive problems. That suggested Reagan’s brain was slipping just a few years into his 1981-1989 tenure; that decline continued. He was diagnosed with Alzheimer’s disease in 1994. Bush showed no linguistic deterioration; he remained mentally sharp throughout his 1989-1993 tenure and beyond.

Sharon Begley answered reader questions about this article on Facebook. Read the conversation here.

  • I can’t believe how many times this article said the decline could be because he’s 70. I know plenty of people 70 and older who can string together a sentence just fine. A small decline is expected for the uneducated and those who don’t stay mentally active as they age, but a change that drastic? Shame on those Drs for not calling it wat it is. That kind of decline is not normal. Now there might be no decline and he could just be acting because that’s what he thinks people want to hear. But that extreme of a decline is not normal aging.

    • You know, even though I think your comment was generated by compassion, there may be some science that would suggest that his age is not the significant factor. I am not up to date, but the general rule was that verbal skills decline little with age and it is the procedural ones that tend to fade, esp. in terms of speed. Actually I.Q. tests did, and I believe still do, address this with different norms for age. So if we are seeing a decline in verbal function, then that may hint that the problem is not ageing, but rather some other neurological degeneration. However, I’m not expert in these matters. I would really like to here from someone who is though.

  • Trump. Well, I haven’t got a single good thing to say about him. I have watched his comings and goings ov TV since he first appeared. I never found him the least inspiring, entertaining, or intelligent. He is very wierd, and getting more so.

    There is senile dementia in some of my family. My mother, for example, died in an Alzheimer’s care facility at age 78. At 71, I worry about myself, and take Aricept, the only FDA approved product for the condition.

    We could have had the first female POTUS in our history. Trump is a national tragedy.

    • Interesting. Also, as a high school teacher, I did notice that lying students also repeated to give theirselves time to come up with a confabulation, besides the ‘uh’, ‘I mean’, Honest, I promise, yuge, yugely.

  • Isn’t it unethical to publicly comment on anyone’s mental status when you aren’t seeing them as a patient? Even then, you wouldn’t be able to disclose without the patient’s permission.

    If Sanders had made it to office, and psychologists started offering commentary on his cognitive process, I would be just as upset. Surely this crosses some kind of ethical line?

    • i don’t believe that speculation like this is unethical when applied to public figures. actually it is unethical and often illegal to speculate to the public if they are under your treatment. When speculation this happens it can be considered rude perhaps, and there may be ethical problems if you are using your expertise for some gain. I believe that when someone holds power at the will of the people, they have rescinded any such right because their mental status threatens others. To me it is more immoral to hold back.

  • What exactly is the point ? Would it be realistic to believe one is the same at 70 that they were at 30-40 ? What a waste of time. Just imagine what could have been accomplished with the time spent on this.
    There is not one of us that will certainly be the same at 70 that we were at 30-40. Isn’t that life ?

    • Howard, the point is, Reagan was later determined to have dementia in his later years in office, and had that been known should have been relieved of duty by the Cabinet and the Vice President installed.

      The question is whether the decline in verbal fluency and reported inability to control temper and so forth are signs of significant cognitive impairment, e.g. dementia. As dementia of all types (FTD particularly, but Alzheimer’s and others well), emotional lability increases and ability to control emotions decreases. If there is even a QUESTION that our President has dementia, he should step down and be replaced. I don’t care WHO it is, Republican OR Democrat. The world is too complex and too many decisions have to be made to take the risk. If, going back in time, the question had actually been ASKED by the Washington press corps about Reagan rather than buried and not mentioned, that might have happened.

  • The better approach would be to quantify verbal fluency, language level, and so forth as applied to verbal comments. This article describes qualitative approaches.

    Those techniques and metrics have been reliably applied to written corpuses (e.g. “grade level” of language, the types of verbs used, and linguistic complexity), but we don’t have those metrics and tools for verbal comments.

    Further, these metrics then need to be tested against verbal fluency declines in dementia patients over time, and quantitative data corpuses are necessary to really draw meaningful conclusions.

    For example, in bipolar patients, new data analysis of spoken language has allowed for the reliable determination of whether a patient is manic, and have reliably predicted the onset of a manic episode. That said, those are pilot studies and based on a small patient population, and need to be replicated and scaled up. These approaches quantify how treatment for bipolar patients is succeeding (or not, as the case may be).

    We need experimentally verified metrics developed from substantial, spoken language corpuses from large numbers of patients with dementia and/or cognitive degeneration over time (preferably 10-20 years). That volume is required in order to make objective, clinically validated measurements for spoken language (spontaneous).

    Without that data, one can make arguments about the extent of the cognitive decline, but what degree of age-related cognitive impairment is substantial enough to cause functional differences CANNOT be quantified. And any arguments that ANY politican is incompetent to execute the duties of their office – even a dog catcher – simply have no weight unless there is quantitative data.

    • While I see the science that you are discussing here, I cannot believe such a strict practice is either necessary or practical in this situation. It would be nice to see such metrics done, but in order for that to happen, the subject either has to submit to it, or is forced to an evaluative procedure. That seems virtually impossible. However, we can judge him on his behavior and only on his behavior. And one of these behaviors is his language. As in the dog catcher, if the dog catcher is seen randomly killing dogs and even if he seems to have psychological factors that may effect a final ruling on him, we can remove him from his job for the safety of the public and deal with the technical diagnosis later. In other words, we do judge competency just on behavior and need to do so. I would also point out that cognitive testing, as you seem to indicate, is actually in a stage where it is often vague as to certain questions, such as competency.

    • Dog catcher analogy is good. The reason I was trying to go that route because the dog catcher’s supporters will find excuses to justify even blatant behavior as killing dogs. Appealing to objective measures, as imperfect as they are, seems like it has more appeal to moderates. Removal of the dog catcher is a political exercise. To the extent that one can prove degradation of objective measures – like linguistic fluency – that is the ONLY way I see to provide cover for that to happen. We are unfortunately past a point where “have you no decency sir” has an effect. The dog catcher will not resign, and only his supporters have the power to remove him.

      The only useful advice I have seen comes from South Korea – they impeached THEIR president a couple of months ago for cronyism behavior far less bad than what’s happened here in the USA. They are, weirdly, the closest analogue to the US political system we have. They even have a Conservative party called the Grand Old Party and have a large conservative / evangelical Christian population in that country.

      Liberals won the election and had majorities in their Parliament / Congres to impeach. And they did.

  • Sharon – knock off the mockery of Alzheimer’s Disease. You don’t do that about Cancer or any other fatal disease. There are Americans really suffering with Alzheimer’s Disease in America, and this type of pseudo-science slander and mockery of a fatal disease is what you can write about. Those who really deal with this problem think such comments are just despicable. Shame.

  • As far as the elitist comment goes. This is a lovely old way to argue, it is called a Ad Hominem argument. Basically just labeling the person you are trying to attack and therefore sidestepping any discussion of the material they are presenting. Calling me this name doesn’t change anything that I said. I never wrote, or even suggested, that we be ruled by the ‘elite,’ whoever that would be. I think that we have a culture, as is shown by your labeling, that does not respect intelligence. I am not suggesting that only the ‘elite’ rule, but that we ask for intelligent statesmen to lead. That we encourage everyone to see intelligence as a positive quality. That we start to deny the worship of the stupid We are currently ruled by a minority with money. This is not democracy. This is not rule of the people. You don’t seem to like the idea of intelligence being a desired quality in our leaders, but you seem willing to have corporations enslave you? You are willing to label me, but you only offer a painful platitude. One of the more remarkable things that you wrote was that the choices are limited. Should they be, or should the average good people have better choices. This last election the parties placed before us two awful choices. One of them tapped into the basic pain and panic that the good people have been feeling for some time and has been continually ignored. Why did we not have a better choice?

    • Well said, Edward. I was fortunate enough to be in elementary school shortly after the leaders of our country realized the implications of the USSR’s launch of Sputnik. Namely, that supremacy – or at least expertise – in mathematics and science is a requirement for our national security. So while my early education in a small town was lacking in the humanities, it was very strong in mathematics and science.
      And while I was labeled a “nerd,” I never encountered anyone who believed mathematics and science were “all lies from the pit of hell,” or boasted about their self-imposed ignorance.

    • Mark,
      Amazing you took a quote about evolution an the big bang by one guy in 2012 “all lies from the pit of hell” and you stretched it to imply Trump and Republucans believe believe that statement about all off science. I complement you as possibly the most dishonest person I have read this week.

Comments are closed.