When Geniuses Nearly Failed Their PhDs
Famous scientists, terrifying oral examinations, and what a viva can—and cannot—tell us
There is a particular kind of academic nightmare that is difficult to explain to anyone who has never experienced a PhD oral examination.
You spend years becoming one of the world's leading experts on a very small subject. You know your experiments, your equations, your literature and your thesis.
Then three professors sit across from you and ask:
"But why did you use that method?"
You answer.
They ask:
"What would happen if you changed that assumption?"
You answer again.
Then, suddenly:
"And what is the resolving power of a microscope?"
And you discover that the entire trajectory of your scientific career may apparently depend on something you last studied as an undergraduate.
It is easy to imagine that the greatest scientists in history would have sailed effortlessly through such examinations.
They did not.
Some performed spectacularly badly. Some were nearly failed. Some were so aggressive, eccentric or intellectually overpowering that the examination became uncomfortable for the examiner.
And a few stories are particularly revealing because the very question that exposed a candidate's weakness later became connected to their greatest scientific achievement.
The most famous example is Werner Heisenberg.
But he is far from the only one.
1. Werner Heisenberg: the PhD oral that almost ended in failure
This is the classic story, and unlike many internet anecdotes about famous scientists, it is exceptionally well documented.
In July 1923, a 21-year-old Werner Heisenberg appeared before the examination committee at the University of Munich.
He was already an extraordinary theoretical physicist.
His supervisor, Arnold Sommerfeld, regarded him as one of his most gifted students. His doctoral dissertation concerned the stability and transition of fluid flow from laminar to turbulent motion—a formidable mathematical problem.
Sommerfeld had deliberately chosen hydrodynamics rather than quantum theory because Heisenberg's interests in the emerging quantum physics were still regarded as unconventional.
The thesis itself was difficult enough that Sommerfeld later said he would not have assigned such a demanding problem to any of his other students.
The thesis passed.
Then came the oral.
And everything went wrong.
The microscope question
One of Heisenberg's examiners was Wilhelm Wien, a Nobel Prize-winning experimental physicist.
Wien was unimpressed by Heisenberg's laboratory skills.
During the examination he asked Heisenberg about the resolving power of a Fabry–Perot interferometer—an instrument Heisenberg had actually encountered in his experimental physics course.
Heisenberg could not derive it.
Wien then moved to something more familiar:
the resolving power of a telescope.
Heisenberg struggled again.
Then:
the resolving power of a microscope.
Again, Heisenberg could not provide the answer.
Wien apparently became increasingly exasperated.
Finally, he asked how a storage battery works.
Heisenberg was unable to give a satisfactory answer to that either.
This was not simply a case of an examiner asking irrelevant trick questions. Wien believed that a physicist should understand experimental physics. Heisenberg had taken Wien's laboratory course and had performed poorly in it.
From Wien's perspective, this was a genuine deficiency.
He wanted to fail him.
Sommerfeld disagrees
Arnold Sommerfeld strongly disagreed.
The committee was effectively confronted with two very different assessments of the same student.
Wien saw an experimental physicist who could not answer elementary questions about optical instruments or a battery.
Sommerfeld saw an extraordinarily gifted theoretical physicist with an exceptional command of mathematics and physical intuition.
The result was a compromise.
Heisenberg received the lowest passing grade in physics and the same overall grade for the doctorate.
He was devastated.
He left a small celebration at Sommerfeld's home, packed his things and took the midnight train to Göttingen.
The next morning he appeared in Max Born's office and asked whether Born still wanted him as an assistant.
Born had him work through the questions he had failed before deciding that Heisenberg's performance was not sufficient reason to withdraw the offer.
The irony gets almost too good
Four years later, the microscope came back.
In 1927, Heisenberg was developing what became the uncertainty principle.
To explain the physical meaning of quantum mechanics, he considered a hypothetical gamma-ray microscope.
The basic idea was that if you wanted to determine an electron's position extremely accurately, you would need radiation with an extremely short wavelength.
But the photon used to locate the electron would also transfer momentum to it.
Better positional resolution therefore came at the price of greater uncertainty in momentum.
The famous relation emerged:
Δx Δp ≥ ℏ/2
The historical irony is extraordinary:
The man who had nearly failed his doctorate because he could not explain the resolving power of a microscope subsequently used a microscope as part of the physical argument surrounding the uncertainty principle.
There is an important qualification, however. The popular version of this story sometimes suggests that Wien's question directly inspired the uncertainty principle. The evidence does not establish such a simple causal connection. Heisenberg's later microscope argument was also technically imperfect and was criticised by Niels Bohr. The modern uncertainty relation is more fundamentally a mathematical property of quantum states than simply a statement about disturbance by an imperfect measuring instrument.
2. David Hestenes: the physicist who actually failed his doctoral oral
Heisenberg nearly failed.
David Hestenes actually did.
Hestenes later became an influential American mathematician and physicist, particularly known for the development and advocacy of geometric algebra, a mathematical framework combining algebra and geometry.
But his path through graduate school included a spectacular setback.
In his autobiographical account of the development of geometric algebra, Hestenes recalls failing his doctoral oral examination.
The committee asked him to solve a standard problem in quantum mechanics.
Unfortunately, it happened to concern one of the gaps in his background.
He struggled.
Eventually he managed to solve the problem after about half an hour, but the committee was unimpressed. They required him to spend another year studying before repeating the examination.
Hestenes later came to regard the decision rather differently.
He thought that, with hindsight, he had actually performed better than many students might have, and that the committee could perhaps have passed him.
But he also concluded that the committee had been trying to protect him from progressing too rapidly.
That is an unusually mature interpretation of failing an oral exam.
The failure was not necessarily saying:
"You are not capable of doing research."
It was saying:
"There is a hole in your foundation, and we think you should fix it before proceeding."
The distinction matters enormously.
3. J. Robert Oppenheimer: when the examiner felt like he was being examined
Not every frightening oral examination ends in a poor grade.
Sometimes the candidate terrifies the examiner instead.
J. Robert Oppenheimer completed his PhD at the University of Göttingen in 1927 under Max Born.
The oral examination became legendary for a very different reason.
After the examination, James Franck—the Nobel Prize-winning physicist who administered the examination—was reportedly heard to say:
"I'm glad that's over."
The reason? Oppenheimer had apparently been so intellectually aggressive and probing that Franck felt that the candidate was getting close to questioning the examiner.
The story should be treated as a reported anecdote rather than a verbatim transcript of the examination. But the general picture fits documented descriptions of Oppenheimer's personality at Göttingen.
He was intellectually intense, exceptionally quick and prone to taking over discussions.
His fellow students eventually became sufficiently frustrated that some reportedly petitioned Max Born to do something about Oppenheimer's dominance in seminars.
Born left the petition where Oppenheimer could see it.
The message worked.
Oppenheimer's case illustrates a different kind of viva problem
A PhD oral examination is not simply an intelligence test.
A candidate can know an enormous amount and still perform badly.
Conversely, a candidate can be so intellectually forceful that the examination becomes difficult to control.
Oppenheimer seems to have belonged to the second category.
His problem was not lack of knowledge.
It was almost the opposite:
He could generate questions faster than the examination could contain them.
That is a very different viva disaster.
4. Hermann J. Muller: the Nobel laureate who was terrified before his oral
The story of geneticist Hermann J. Muller gives us a more human version of the same problem.
Muller would eventually win the 1946 Nobel Prize in Physiology or Medicine for his discovery of X-ray-induced mutations in Drosophila.
But before his doctoral examination at Columbia University in 1915, he was extremely nervous.
His mentor Edmund Beecher Wilson tried to calm him by telling him about his own nightmare before his doctoral examination decades earlier.
Wilson described dreaming that he was standing before his examiners—and that among them were Charles Darwin, Thomas Huxley and Ernst Haeckel.
Wilson announced that he had solved the problem of life.
He then drew two triangles on the blackboard.
His conclusion?
That was the secret of life.
In the dream, Darwin threw his hat into the air and congratulated him.
Muller later remembered the story.
The remarkable thing is that this is not merely an internet anecdote. The Cold Spring Harbor Laboratory Archives reconstructed the episode from archival material, including the diary of geneticist and historian Elof Carlson, who had heard the story directly from Muller.
Muller did not fail his oral.
But the story is worth including because it shows something that is often forgotten when we look backward at Nobel laureates:
They were once terrified graduate students too.
5. Hugh Huxley: a thesis examination at the edge of a scientific revolution
Another fascinating biology example involves Hugh Huxley, one of the founders of modern muscle biology.
Huxley was working on the fine structure of muscle at a time when the interpretation of muscle structure was undergoing a major transformation.
Researchers were beginning to realise that muscle contraction could not be explained simply by the shortening of individual protein filaments.
The emerging idea was that two sets of filaments might instead slide past one another.
Around this period, Huxley's work, together with the work of Andrew Huxley, John Kendrew, Francis Crick and others, helped establish the structural basis for what became the sliding-filament model of muscle contraction.
The important point for our discussion is that a PhD examination can take place at precisely the moment when the candidate is still figuring out what the evidence means.
The oral examination is therefore not necessarily the interrogation of a finished scientist.
It is an assessment of someone who is still becoming a scientist.
6. Sydney Brenner: when the examiner nearly failed the candidate because he disliked the thesis
Sydney Brenner provides perhaps the most entertaining perspective because he was himself an examiner.
In his recollections, Brenner describes examining a PhD thesis in biochemistry.
He was unimpressed by the dissertation.
The work, in his view, contained speculative sections unsupported by direct experimental evidence and repetitive experiments. He drafted an extraordinarily sarcastic report recommending that the candidate not receive the degree.
Then he reconsidered.
Brenner realised that he had no objective basis for allowing his personal irritation with the thesis's author to determine the candidate's fate.
So he discarded the devastating report and wrote a conventional one.
The candidate received the PhD.
The story is valuable precisely because it comes from the examiner's side.
A viva can be influenced by something surprisingly dangerous:
the examiner's personality.
An examiner can become annoyed by:
- the writing;
- the supervisor;
- the choice of terminology;
- the interpretation of the data;
- the field itself;
- or simply the candidate's manner.
The scientific merits of the thesis can become entangled with these reactions.
Brenner recognised that danger in himself.
And he corrected for it.
7. Philip Eaton: when the chemistry examiners themselves disagreed
Chemistry provides fewer famous examples of outright PhD-viva disasters, at least ones that are documented well enough to survive historical scrutiny.
But there are some wonderful examination stories.
One comes from organic chemist Philip Eaton, whose PhD research involved early applications of NMR spectroscopy.
During his oral examination, two giants of organic chemistry—John D. Roberts and Robert Burns Woodward—became involved in a vigorous argument over the interpretation of Eaton's spectra.
Eaton later realised that both examiners had been wrong about the interpretation.
This is a beautiful example of why a PhD oral is not necessarily an interrogation in which the examiner possesses the answer and the candidate is supposed to reproduce it.
Sometimes the candidate is closer to the truth than the examiner.
And sometimes the examiner's expertise is itself bounded by the technology and conceptual framework of the time.
NMR was still developing rapidly.
What seems obvious in a modern textbook can be genuinely uncertain at the frontier of a field.
8. Linus Pauling: a reminder that even great scientists have historical blind spots
Linus Pauling is an interesting chemistry counterexample.
There is no strong evidence that Pauling nearly failed his PhD oral. He successfully completed his PhD at Caltech in 1925.
But his examination is interesting from a historical perspective because the chemistry of his doctoral period was undergoing an enormous conceptual transformation.
A modern Pauling transported back to his doctoral examination with decades of subsequent knowledge would know quantum chemistry, spectroscopy and laboratory technologies that his examiners could scarcely have imagined.
This raises an important question:
How much of a PhD oral should test timeless knowledge, and how much should test the candidate's mastery of the scientific world available at the time?
That question is particularly important in rapidly changing fields.
9. The famous oral examinations that probably didn't happen the way the internet says
There is another lesson in researching these stories.
The internet is full of claims such as:
- "Einstein failed his PhD oral."
- "A Nobel laureate couldn't answer basic questions."
- "The examiner tried to fail a famous scientist."
- "A professor asked a trick question that changed the history of science."
Many of these stories are either exaggerated or involve a different examination altogether.
For example, one famous story concerns Mileva Marić, Einstein's future wife.
She failed the final ETH examination in 1900, whereas Einstein passed narrowly. But this was not a PhD oral examination. It was the final examination for the ETH diploma.
Similarly, many American scientists underwent qualifying examinations or comprehensive examinations that should not be confused with the final oral defence of a doctoral thesis.
This distinction matters.
A PhD student can fail:
- a qualifying examination;
- a comprehensive examination;
- a candidacy examination;
- a thesis proposal defence;
- a final thesis defence;
- or a separate oral examination in a particular subject.
These are not interchangeable.
Why are genuinely failed famous PhD vivas so rare?
There is an interesting historical reason.
If a truly famous scientist had an ordinary PhD oral, historians often simply never recorded it.
But if the examination was disastrous, humiliating or bizarre, it was much more likely to become part of the scientist's biography.
This creates a strong selection effect.
We remember:
Heisenberg could not answer the microscope question.
We do not remember the thousands of future professors who gave perfectly competent answers to every question.
There is another reason.
Most universities do not want the final PhD examination to function as a lottery.
By the time someone reaches a thesis defence, the supervisor, department and university have already invested years in the candidate.
The thesis has been submitted.
The examiners have been selected.
The candidate's research record is available.
A complete failure should therefore normally reflect something substantial:
- the thesis does not constitute an original contribution;
- the candidate cannot explain or defend central results;
- the data do not support the conclusions;
- there are major methodological flaws;
- the candidate does not understand the work;
- or the candidate cannot demonstrate the expected level of independent scholarship.
A candidate simply forgetting a formula is not necessarily enough.
The most frightening question is often not the hardest question
This may be the biggest lesson from these historical cases.
The question that causes a candidate the most trouble is not necessarily the one requiring the most advanced knowledge.
Heisenberg's problem was not an obscure equation in quantum field theory.
It was:
How does a microscope resolve two nearby objects?
For another candidate it might be:
Why did you choose this statistical test?
Or:
What is the parameter you defined in Chapter 4?
Or:
What would happen if this assumption were violated?
These questions are frightening because they probe whether the candidate actually understands the foundations of their own work.
An examiner who asks a difficult but relevant question may discover that the candidate knows the subject extremely well.
An examiner who asks a seemingly simple question can sometimes discover something much more serious:
The candidate has been operating a method without understanding it.
The Heisenberg lesson has a dangerous corollary
It would be easy to draw the wrong conclusion from these stories.
The lesson is not:
"Oral examinations are useless because Heisenberg almost failed one."
That would be absurd.
Wien identified a genuine weakness in Heisenberg.
Heisenberg really did have poor experimental knowledge.
The lesson is instead:
A PhD examination samples a multidimensional ability using a very small number of observations.
Heisenberg's oral revealed something true about him.
It simply did not reveal everything that was true about him.
That distinction is crucial.
What should a PhD oral actually determine?
A good doctoral examination should ideally establish several things.
1. Does the candidate understand the thesis?
Not just the conclusions.
The candidate should understand:
- why the question was asked;
- why particular methods were chosen;
- what the assumptions were;
- what the alternatives were;
- what the limitations are;
- and what the results actually demonstrate.
2. Can the candidate reason beyond the thesis?
This is where a good viva becomes much more than a presentation.
The examiner should be able to ask:
"What if...?"
and see whether the candidate can reason through the consequences.
3. Does the candidate understand the foundations?
A candidate doing evolutionary genomics should know more than the commands used to run an analysis.
A candidate doing molecular biology should understand more than the protocol.
A candidate doing theoretical physics should understand more than the equations in the thesis.
A doctorate represents independent scholarship, not merely successful completion of a research project.
4. Can the candidate recognise uncertainty?
This may be one of the most important signs of scientific maturity.
The strongest candidates can say:
"I don't know."
and then continue:
"But here is how I would find out."
That is much more impressive than inventing an answer.
And what should an examiner not conclude?
A candidate who cannot answer one question is not necessarily incompetent.
A candidate who becomes nervous is not necessarily unprepared.
A candidate who forgets a formula is not necessarily ignorant.
And a candidate who disagrees with an examiner is not necessarily wrong.
The history of science contains plenty of cases where examiners themselves were wrong.
The Philip Eaton story is a useful reminder of this.
The Heisenberg story is another.
The truly dangerous viva
The most interesting cases are therefore not necessarily those in which the candidate says:
"I don't know."
The genuinely dangerous cases are those in which the candidate repeatedly demonstrates that they do not understand what they have done.
There is a profound difference between:
"I did not consider that possibility."
and:
"I don't understand why this analysis was done."
There is also a difference between:
"I don't remember the exact derivation."
and:
"I cannot explain what this parameter means."
A PhD oral should be able to distinguish those situations.
That is much more difficult—and much more meaningful—than simply trying to catch a candidate out.
What these famous disasters ultimately tell us
Heisenberg's examination tells us that a future revolutionary can have a glaring weakness.
Hestenes's failed oral tells us that failure can sometimes be a useful intervention rather than a final judgement.
Oppenheimer's examination reminds us that intellectual intensity can be as challenging for an examiner as ignorance.
Muller's story reminds us that even Nobel laureates once stood outside examination rooms terrified.
Huxley's story shows that important scientific ideas can emerge while a scientist is still in training.
Brenner's story reminds examiners that their own emotions can influence judgement.
And Eaton's chemistry examination reminds us of something even more fundamental:
The examiner is not necessarily right.
Science is not a ritual in which an all-knowing professor interrogates an ignorant student.
At its best, a PhD oral is a meeting between scientists in which one side is testing whether the other has genuinely crossed the threshold from student to independent researcher.
Sometimes that process goes badly.
Sometimes spectacularly badly.
Sometimes the candidate deserves to fail.
Sometimes the examiner gets it wrong.
And occasionally, as in Heisenberg's case, a candidate can leave the examination room with the lowest passing grade—and go on to transform an entire field of science.
The most comforting lesson for a nervous PhD student is therefore not:
"Don't worry. Even geniuses fail exams."
It is something more useful:
An oral examination can reveal what you know, what you don't know, and how you think under pressure. But it is not a crystal ball.
A bad viva can be a serious warning.
It can be a deserved failure.
It can be a temporary setback.
Or, occasionally, it can be one afternoon in which an extraordinarily uneven human being happens to be asked exactly the questions he is least equipped to answer.
And that, as Werner Heisenberg discovered, is not necessarily the end of the story.
Sources and further reading
- American Physical Society — historical account of Werner Heisenberg's doctoral oral examination.
- American Institute of Physics — The Sad Story of Heisenberg's Doctoral Oral Exam.
- Cold Spring Harbor Laboratory Archives — historical material concerning Hermann J. Muller and Edmund Beecher Wilson.
- Historical accounts and autobiographical writings of David Hestenes concerning his doctoral oral examination.
- Biographical accounts of J. Robert Oppenheimer's time at Göttingen.
- Historical accounts of Philip Eaton and the early development of NMR spectroscopy.
Note: Stories about famous scientists and examinations are particularly vulnerable to embellishment. Where an anecdote is not supported by a contemporary examination record or a reliable retrospective source, it should be treated as a reported story rather than as an exact transcript of what happened.