Artificial Intelligence

Throughlines of Genius

Al-Kindi and the origins of machine translation

By Cameron Rasmusson

W

hen Alan Turing set his mind to decrypting Nazi codes at the advent of World War II, did he consciously recognize his place in a millennium of scientific development?

What about the humble translators who embraced machine translation (MT) during its expansion a half-century later in the 1980s and ‘90s? When they sat down at their MS-DOS or Apple computers to fire up the latest software in their computer-assisted translation (CAT) systems, did they see themselves as inheritors of an intellectual throughline tracing back to the ancient Greeks or the Abbasids of the Middle Ages?

Whether they knew it or not, and whether history remembers their names or not, all these individuals were players in a story almost as old as human remembrance itself. And while it’s a sprawling story with a huge — almost overwhelming — cast, modern scholarship traces its origins back to one individual in particular: the Arab Muslim philosopher, mathematician, physician, and music theorist Abu Yusuf al-Kindi.

Advertisement

“One Must Not Be Afraid of New Ideas”

When it comes to truly intimidating résumés, it’s hard to top al-Kindi. A polymath through and through, he wielded the kind of insatiable intellectual curiosity characteristic of history’s greatest scientific achievers, from Isaac Newton to Albert Einstein. According to Henry Corbin’s History of Islamic Philosophy, he wrote 32 books on geometry, 22 on medicine and philosophy, nine on logic, and 12 on physics — producing at least 260 books over a lifetime’s body of work.

A member of the Arab tribe of the Kinda, al-Kindi enjoyed a privileged upbringing as a member of his tribal aristocracy — his father being the governor of Basra — with the era’s best educational resources at his disposal. According to Corbin, he then went on to complete his studies in Baghdad, where, recognizing his talents, his educators appointed him to the House of Wisdom, a crucible of Middle Ages-era learning. Established in the 8th century, the House of Wisdom was a thriving intellectual bastion by al-Kindi’s lifetime, estimated to be between 801 and 873 CE. Access to these elite institutions is likely what exposed the budding scholar to the works of the ancient Greek philosophers — a key link in an intellectual chain that persists even today.

Al-Kindi’s surviving writings paint a portrait of a man with endless curiosity and an intellectual open-mindedness remarkable for any historic era. “One must not be afraid of new ideas, no matter the source; and we must never fear the truth, even when it pains us,” reads one popular al-Kindi quote.

“We ought not to be embarrassed of appreciating the truth and of obtaining it wherever it comes from, even if it comes from races distant and nations different from us,” reads another. “Nothing should be dearer to the seeker of truth than the truth itself, and there is no deterioration of the truth, nor belittling either of one who speaks it or conveys it.”

To be sure, it’s an open-mindedness that sometimes landed al-Kindi into trouble. Corbin suggests that his unorthodox perspectives were often disfavored by the academic and civic authorities of the time, leading to a lonely death in Baghdad by the late 9th century. And according to Peter Adamson, professor of late ancient and Arabic philosophy at Ludwig Maximilian University of Munich, he was at one point beaten and his library confiscated.

Al-Kindi’s at-times bitter life resulted in a perspective about personal attachments not dissimilar to Robert De Niro’s character Neil McCauley in the classic heist film Heat: “Don’t let yourself get attached to anything you are not willing to walk out on in 30 seconds flat if you feel the heat around the corner.” The famous movie line is almost a direct echo of al-Kindi’s own words; “the lesson to draw may not be that we should avoid forming attachment entirely, but that we should do so sparingly and only for good reason,” the scholar wrote.

Perhaps due to his frequent unpopularity in his own age, most of al-Kindi’s work has been lost to time. According to Felix Klein-Franke’s 2001 book Al-Kindi, what surviving works we possess can largely be credited to Gerard of Cremona, a 12th-century Italian translator from Arabic into Latin and a central figure in the famed Toledo School of Translators.

Indeed, translation is a binding theme throughout the life of al-Kindi. The House of Wisdom, which shaped and disciplined his mind, was itself a center for the translation of ancient Greek literature into Arabic. His own work, or what survives of it, is available thanks to Italian translators. And translators of the modern era still benefit from al-Kindi’s ideas in ways he could never have imagined.

Figure 1. The first page of al-Kindi’s manuscript Treatise on Deciphering Cryptographic Messages, which proved the Arabic origins of frequency analysis.

“Count the Occurrences of Each Letter”

As Quinn DuPont observes in his article “The Cryptological Origins of Machine Translation,” there’s a fundamental link between cryptology and translation. After all, both are fundamentally about decoding an indecipherable text, whether by design or by language barrier, and making its meaning plain. But prior to academic discoveries in the 1980s, the prevailing historic narrative pointed to an origin of MT in the 16th and 17th centuries. More recent scholarship challenges that notion.

“I complicate and extend this narrative of machine translation by tracing it through a longer history of cryptanalysis. This history begins before the Western precursors in the sixteenth and seventeenth centuries, and distinguishes itself from these latter efforts, which were largely cryptographic (involving encryption/decryption), and not necessarily cryptanalytic,” DuPont writes.

“During the so-called Islamic ‘Golden Age’ (a somewhat problematic catch-all term, denoting a period of intellectual vibrancy from roughly the eighth to thirteenth centuries of the common era), Arabic scholars possessed the suitable intellectual and political skills and the necessary motivations for the development of cryptanalysis,” DuPont continues later in his article.

Citing the late David Kahn, perhaps the most respected scholar on the history of cryptography, DuPont suggests that the nature of the Arabic political state, the evolution of its bureaucracy, the sophistication of its mathematics, and the cultural reverence for language, among other factors, all contributed its influence on translation and cryptography. According to Mohammed I. Al-Suwaiyel of King Abdulaziz City for Science and Technology, Kahn based his assertion on references within historical documents to a book by 14th-century scholar Ali ibn ad-Durayhim: Miftah A-Kunuz fi Idah Al-Marmuz, or Key to Treasures on Clarifying Ciphers.   

Figure 2. Arabic artwork depicting Aristotle, who, along with other ancient Greek scientists and philosophers, proved influential on Middle Age Arabic scholarship.

But can you take a reference in a historic manuscript at face value? That was the question before 1980, and it kicked off an international scavenger hunt.

“In 1979, Drs. M. Mrayati, Y. Alam, and M. al-Tayyan, from the Arab Academy of Damascus, decided to verify the truth of Kahn’s statement and look for ibn ad-Durayhim’s lost book,” Al-Suwaiyel said in a 2018 presentation. “Dr. Mrayati and his team discovered a treasure! Not only [did they find] ibn ad-Durayhim’s book, but they also discovered more than 15 Arabic manuscripts on cryptology written by Arab scholars in the … 9th to 15th centuries AD.”

Among that treasure trove of historical texts was al-Kindi’s Treatise on Deciphering Cryptographic Messages. And suddenly, a new world of science history revealed itself.

“[It’s] the oldest extant manuscript on cryptanalysis written in the 9th century AD,” Al-Suwaiyel said in his presentation. “The manuscript is about 1,200 years old!”

What al-Kindi pioneered in his treatise was essentially an early version of the code-breaking technique called frequency analysis, among other principles. He described the process as follows: “One way to solve an encrypted message, if we know its language, is to find a different plaintext of the same language long enough to fill one sheet or so, and then we count the occurrences of each letter. We call the most frequently occurring letter the ‘first,’ the next most occurring letter the ‘second,’ the following most occurring letter the ‘third,’ and so on, until we account for all the different letters in the plaintext sample.”

Figure 3. Al-Kindi depicted on an Iraqi postage stamp in 1962.

“Then we look at the ciphertext we want to solve, and we also classify its symbols,” al-Kindi continued. “We find the most occurring symbol and change it to the form of the ‘first’ letter of the plaintext sample, the next most common symbol is changed to the form of the ‘second’ letter, and the following most common symbol is changed to the form of the ‘third’ letter, and so on, until we account for all symbols of the cryptogram we want to solve.”

By analyzing the frequency of letters in an Arabic text of 3,667 letters, he established a framework for identifying patterns and decoding texts, thus laying the groundwork for both cryptography and MT.

“Al-Kindi noted how common words or phrases, such as the traditional opening for Islamic works, ‘In the name of God, the Compassionate, the Merciful,’ could be used as a guide throughout the cryptogram,” DuPont wrote. “That is, if a cryptanalyst could find ciphertext corresponding to certain common or probable words or phrases, guesses could be made to reveal the corresponding plaintext letters, which may also reveal a key or common substitution used throughout.”

Advertisement

“Those Who Have Contributed Much Truth”

It’s difficult to overstate how the principles laid out by al-Kindi in the 9th century have shaped life today. Cryptanalytic systems and technologies underpin our financial transactions, internet use, communications, and, of course, the translation systems that professional linguists rely upon every day. And the thought that its origins can be traced back to inquiring minds who lived and worked over 1,200 years ago is extraordinary.

From our limited human perspective, it’s impossible to understand how our thoughts and actions will ripple beyond our mortal constraints. That’s a truly humbling notion, albeit one that supplies valuable perspective and clarity. If nothing else, we can at least look back with gratitude at the bright lights who set the foundations of the modern world.

And that’s a perspective al-Kindi himself would likely embrace. He found inspiration and insight in the ancient Greek thinkers he studied, an intellectual debt he acknowledged in his own work.

“It is proper that our gratitude be great to those who have contributed even a little of the truth, let alone to those who have contributed much truth, since they have shared with us the fruits of their thought, and facilitated for us the true (yet) hidden inquiries, in that they benefited us by those premises which facilitated our approaches to the truth,” al-Kindi wrote. “If they had not lived, these true principles with which we have been educated, towards the conclusions of our hidden inquiries, would not have been assembled for us, even with intensive research throughout our time.”

CAMERON RASMUSSON is a senior writer and editor for MultiLingual Media.

Advertisement

Related Articles