I recently watched a TED talk with a fun topic: why is ‘x’ used as the unknown variable in Algebra and beyond? If you haven’t seen it, this 4-min talk is above. The thesis is this: x is used, because the Spanish, who were the first to translate the great scientific and engineering knowledge coming from the Arab world, don’t have in their language the ‘sh’ sound that dominated the original Arabic word, ‘something‘. Hence, by convention, they used the sound ‘k’ as in the Greek letter chi (χ) (pronounced ‘kah-i’ in English); later translations of these works to Latin mapped that to the Latin letter x, because the symbols look the same. The speaker is engaging and confidant, the talk is short and sweet, it’s a great story to tell over a dinner party. Except for one detail: it’s full of historical inaccuracies.
Statement: “Spanish doesn’t have the ‘sh’ sound” (@ 2:53)
Historical inaccuracy #1: The statement is true-ish for modern Spanish. But at the time that the Moors were in the Iberian Peninsula, the languages spoken there were quite different from modern Spanish. The one directly related to modern Spanish is called Old Spanish, aka Castilian Spanish. Scholars believe that Old Spanish had the sound ‘sh’. One the victims of evolution of the Spanish language was precisely that sound; in modern Spanish it has been replaced by the softer form ‘hj’ (as it sounds in the German word ‘ich’). Modern Spanish also has the ‘tsh’ sound, like “Muchos muchachos.” So while it is true that “[modern] Spanish doesn’t have the ‘sh’ sound,” it does have the softer form of that sound and the form with the preceding ‘t'; and back then, when the works were supposed to have been translated, it is believed by Scholars to have had the ‘sh’ sound.
Historical inaccuracy #2: The statement makes one believe that Spanish is the language of the Iberian Peninsula, and the only one that matters for this issue. At the time the Moors were in the Iberian Peninsula, many languages were spoken there, not just one. And at the time the translations were made, Arabic was still pervasive, especially in the intellectual centers where translation of these works is believed to have happened (Toledo). Even now, there isn’t just Spanish, but Portuguese, Galician, Catalan, Basque, along with many more dialects of these. All these languages had the sound ‘sh’, and most of them, most notably Portuguese, still have it pervasively. (Portuguese sounds like Russian) So the ‘sh’ sound was far from being a foreign, difficult phoneme to pronounce and translate anywhere in the Iberian Peninsula.
Statement: “By convention, they borrowed the ‘k’ sound from the classical Greek in the form of the letter chi (χ)” (@ 3:00)
Historical inaccuracy #3: Only English-speaking people pronounce the chi letter as ‘kah-i’. Everyone else, especially the Greeks, pronounce it ‘jhee‘ — the ‘jh’ as in ‘ch’ of the German word ‘ich’. This is the softer version of the ‘sh’ phoneme. There doesn’t seem to be the need for a convention going from the word ‘something’ in Arabic to using the Greek letter χ to capture its first phoneme. (Follow the links to hear the sounds) Phonetically, they are very close. According to Wikipedia, ‘something’ in Arabic was translated to ‘χei’, which, if you pronounce the χ correctly like the Greeks do, sounds almost identical.
No mystery here. The use of the then-new word ‘χei’ seems to be an instance of the frequently-occurring phenomenon of appropriating foreign words phonetically into a host language and using the symbols of the host language for its written form. (E.g. coffee in Japanese)
Statement: “Later, when this material was translated [...], they simply replaced the letter χ with the Latin X.” (@ 3:11) [the slide showing χ morphs into a slide showing X]
Historical inaccuracy #4: It’s unclear who “they” refers to, but the speaker makes it sound like “they” means the scholars who were tasked with translating the Algebra works from “Spanish” to Latin. According to the speaker, when faced with χ in the “Spanish” version, these scholars simply replaced it with X, as if by accident — because the written symbol was similar. This is not accurate: the Latin letter X evolved from the Greek letter χ, that’s its root. Its use in these works was not an historical accident, it’s essentially the same letter that denotes the phoneme ‘jh’.
In fact, in Portuguese we still pronounce X as ‘sheej‘, very close to the Greek letter χ, which, in turn, is phonetically similar to the first phoneme of the word ‘something’ in Arabic. (Again follow the links to hear the sounds)
Additional historical reference: Original Iberian texts (i.e. not translations, but actual original works) that have been preserved from that time show the use of the letter χ extensively. Its use denotes the fricative phoneme ‘jh’ or ‘sh’. Here is one concrete example, a Medieval cantiga from the Galaico-Portuguese Cancioneiro. Click the images on the right of that page to see the original script, and click the link on the bottom right of that page to hear the song. (I should disclose that the main reason why I know about this Cancioneiro site is because it has been developed by a team led by my sister Graça)
OK, so this one person steps up to a TED stage and tells an interesting, plausible story [in English] that involves historical accidents (always fun to hear those) and that makes the [English-speaking] audience laugh and applaud. Never mind that this person doesn’t give one single scholarly reference to his thesis. Surely enough, a few days after the talk was published on the Web, someone edited the Wikipedia entry on the letter X adding the thesis of this talk as historical fact. (That edit was removed a few days later)
The Web is a wonderful thing, and conferences like TED that are looking for good storytellers are a wonderful thing too. But I sure am glad to be part of this Medieval community called Academia that not just hears the presentation of plausible theses and good stories, but that forces the presenters to defend those theses with concrete data and solid arguments in front of a group of informed, professional skeptics like me.