In 1883, Alcalá’s work was edited and published by Paul de Lagarde—an “orientalist”, not a mathematician. Evidently Lagarde was aware that Arab mathematicians used that word for the unknown quantity in algebra, and in 1884 he published a speculation that “x” in algebra might have been an abbreviation of the Old Spanish transcription of the Arabic word. That charming theory caught on. Evidently Lagarde was not aware that Spanish mathematicians never used a _transcription_ of the Arabic word—instead, they used the _translation_ in their own language, “cosa”.

Today most historians of mathematics agree that Descartes originated the use of “x” arbitrarily, and first published it in 1637. They would have to revise that belief if an earlier published instance came to light; but so far, no such evidence has been found. ]]>

Which is why Mr Moore is a hit on TED and all the rest of you are not. He in interesting, keep it simple and, in context, uncommonly accurate.

]]>Dear Math Guru and Sean,

Basically the letter x originated from persian. It originates from Omar Khayyam’s work on cubic equations. For the unknown variable Khayyam used the word “shay” which means “thing”.

The andalusian ommayads (basically muslims in Spain.) this word was written as “xay” with the spanish alphabet. After a while only the first letter of this word (x) was used to represent the unknown in equations. Therefore use of x as the unknown became popular. The “y” and “z” came after that.

Considering Khayyam lived around 1100’s I am pretty sure Descartes and mathematicians before him were already using x.

On a further note the word Algebra also originates from a persian mathematician Abdallāh Muḥammad ibn Mūsā al-Khwārizmī’s book “Kitab al-jabr wa al-muqabalah”. Al-jabr roughly translates to “the synthesis”

]]>Sad… but all you conversation comes from beautiful legend and misinformation.

]]>The list of discontinuities should be sufficient for nice functions (i.e., “elementary” functions aka closed form expressions involving polynomials, trig, log, etc.).

]]>I ran across this while scooping UOIT’s Math grad website.

I’m not so convinced about the -N traslation of the function f and it’s purpose. The IVT is basically equivalent to Dedekind completeness (or LUB property ) of the reals. Your reasoning “work” for over the rationals too (because in a sense, they look “complete” when you graph functions over the rationals). And the -N seems unnecessary too.

Thanks

]]>