On 8 April 1982, an Israeli researcher named Dan Shechtman was observing an aluminium-manganese alloy he had rapidly cooled through an electron microscope. What he saw should not have existed: the diffraction pattern, the footprint left by ordered atoms when struck by an electron beam, possessed a tenfold symmetry: it rotated upon itself ten times in a complete revolution and aligned with itself each time. To any crystallographer of the era, this was akin to photographing a four-sided triangle. In his notebook, Shechtman recorded three words that summarise a century of wavering certainty: '10 fold ???', followed by three question marks.
One must understand why it was considered impossible. Crystallography had been constructed upon a seemingly unbreakable law: crystalline matter is that which arranges its atoms periodically, repeating the same motif over and over, much like wallpaper or a tiled floor. Geometry demonstrates that if one wishes to fill space by repeating a motif without leaving gaps, it can only be achieved with certain symmetries: of order two, three, four, or six. Never of order five, never of order ten. Just as one cannot tile a floor using only pentagons without leaving gaps. The symmetry Shechtman was observing was, by definition, incompatible with being a crystal. He spent hours verifying whether it was a deception, two crystals joined together to imitate that symmetry. It was not.
The discovery did not elicit indifference, but rather the active defence of a definition. The head of his research group brought him a crystallography textbook and suggested he read it, so that he might understand that what he claimed to have seen could not exist. Ultimately, he requested that he leave the group. Years later, when asked openly if he had been expelled from the laboratory, Shechtman replied, without drama, that he had. He was not dismissed for being mistaken, but for insisting upon a datum that contradicted what everyone knew to be true.
Shechtman was not alone. John Cahn, an eminent chemist, initially dismissed him: 'Go away, Danny, these are two crystals stuck together and it is of no interest', but shortly thereafter he changed his stance and issued a challenge that proved decisive: 'this material is trying to tell us something, and I challenge you to find out what it is'. Ilan Blech constructed the model that transformed the observation into something publishable. Others co-authored the article that was finally published in 1984, two and a half years after that morning. It was not one against all: it was uncomfortable evidence making its way, with allies, against a largely hostile community.
The most illustrious adversary arrived later and was formidable: Linus Pauling, a two-time Nobel laureate and the greatest living authority on chemical bonds. Pauling refused to accept quasicrystals until the end, maintaining in articles that what Shechtman observed were, once again, crystals joined in a complex manner. A devastating phrase is attributed to him: 'there are no quasicrystals, only quasi-scientists', which circulates widely, although I have not found the primary source to certify it, so I leave it as what it is: an attributed remark. What is documented, however, is his signed and tenacious opposition. The highest authority in the field exerted all his influence to deny the datum.
And yet. In 2011, nearly thirty years after that morning, Shechtman received the Nobel Prize in Chemistry for the discovery of quasicrystals. The community that had expelled him had to perform a task more difficult than admitting a specific error: it had to change its own definition of what a crystal is. The International Union of Crystallography rewrote it to include forms of order that are not periodic. What had appeared to be a violation of the laws proved to be an expansion of what those laws described.
I do not draw a moral here against science, but rather on how it functions. Shechtman did not demonstrate that crystallography was incorrect; he demonstrated that its definition of a crystal was narrower than physical reality, that there were possible arrangements which the definition did not contemplate and therefore declared impossible. The problem was never with the atoms: it was in the phrase we had decided upon to define what constituted a crystal. When well-verified data collides head-on with a definition, there are two outcomes: to deny the data or to expand the definition; for years, the field chose the former. Science eventually chose the latter, which is the correct path, but it took nearly three decades and cost a man his place in the laboratory. Definitions serve us for thinking, until the day reality presents something that does not fit within them. On that day, courage does not consist of defending the definition: it consists of believing what one sees.
On the open conversation
This text recounts a verified episode in the history of science, the discovery of quasicrystals by Dan Shechtman in 1982 and his prolonged rejection until the 2011 Nobel Prize, to reflect upon what occurs when verified data does not fit within the prevailing definition of a field. I distinguish the fact (the observation, the documented rejection, the change in definition) from the legend (the solitary genius, which he was not: he had early allies) and mark the celebrated phrase by Pauling as attributed, not as a firm citation. If anyone wishes to intervene from the perspectives of physics, the history of science, or epistemology, this notebook remains open.
Sources
NIST, 'The Nobel Moment: Dan Shechtman' (2016): the observation of 8 April 1982, the notation '10 fold ???' and the challenge from John Cahn.
Nobel Foundation: 2011 Nobel Prize in Chemistry 'for the discovery of quasicrystals'; interview with Shechtman (2011) confirming his expulsion from the group.
Shechtman, Blech, Gratias and Cahn, 'Metallic Phase with Long-Range Orientational Order and No Translational Symmetry', Physical Review Letters (12 November 1984).
Linus Pauling, 'So-called icosahedral and decagonal quasicrystals are twins of an 820-atom cubic crystal', Physical Review Letters (26 January 1987). The phrase 'there are no quasicrystals, only quasi-scientists' is widely attributed to him, without a primary source located.
International Union of Crystallography: expansion of the definition of a crystal (1991 report, published in 1992).