BAD ETYMOLOGY. I'm used to seeing dubious or just plain wrong etymologies, both online and off-, and usually I just ignore them.
This site, however, is so bad that I feel the need to give it a public thrashing. It purports to list borrowed words by their languages of origin (and it's the number one Google hit for "borrowed words," so I'm not just picking on some obscure site no one will ever see). Let's take Akkadian, for which the entry is:
Babel (ancient city - from babul, gate of God), Babylon (ancient capital city - from babul, gate of God), cherub (gracious), dragoman (interpreter), Orion (constellation - from Uru Anna, light of heaven), ziggurat
Now, ziggurat is an Akkadian word; no problem there. Babel and Babylon are also from Akkadian, but the etymon is wrong. Here's the best description I've found on the early development of the name: "The Sumerian name for this small village was Ka-dingir-ra. In Semitic Akkadian it was called Bab-ilim. It seems that the name came not from Kadingirra, but from another name for the town, Babil, the meaning of which is unknown. Later the plural name Bab-Ilani 'the Gate of the Gods' was used." It's normal (if oversimplified) to say that Babylon is from Akkadian Bab-il/ilu/ili/ilani (take your pick) and define this as 'gate of God,' but "babul" seems to have been pulled out of a hat (perhaps by vague association with Kabul). The other entries are sheer fantasy. "Cherub" is, of course, from Hebrew. "Dragoman" is (via Italian and Greek) from Arabic
tarjuman, which is from Aramaic
turgemana. And Orion, as any schoolboy knows, is a Greek mythological figure; this etymology is probably taken from the
Online Etymology Dictionary, which says "perhaps from Akkadian
Uru-anna 'the Light of Heaven,'" but that's a very big "perhaps"—what a Greek hunter would be doing with an Akkadian name is anybody's guess, and "etymology unknown" is the only safe statement.
The Afrikaans entry lists "slim," which is from Dutch or Low German. The Albanian entry reads, in its entirety, "Carpathian (Eastern European Mountain Range - from karpë, rock)," which is ridiculous; "Carpathian" is from Greek, and while the Greek name may well have come from a local Thracian or Illyrian word related to Albanian
karpë, that's like saying "Caucasus" is from "high" because there's a possibility the name is related to the Germanic root of the English word. Under Algonquin is listed "Oregon," which is of very disputed etymology; one theory is that it resulted from a French map engraver's having put the last four letters of "Ouariconsint" (the Wisconsin River) on a separate line, thus creating an apparent "Ouaricon" River, and "Wisconsin" is probably from an Algonquin name, but that's really pushing it. The next language listed is "American English," and I won't bother going through the words, because the whole category makes no sense—even if "raincoat" and "typewriter" were first used in the U.S., how can they be considered "borrowed"? An almost equally pointless category is "Anglo-Saxon" (more properly called Old English); most of the basic English vocabulary is "from" Old English in the sense of having developed from it by sound change, but the only words that could be said to be borrowed from it are scholarly ones like "witenagemot," and none of them are listed. Under Amoy is listed "ketchup," which is from Malay; under "Avetsan" (i.e., Avestan) is "bronze," which is from Italian (via French); under Basque is "bizarre," also from Italian; the sole entry for Beja is "bedouin," which is from Arabic; and under Breton is a whole raft of words, none of which have anything to do with Breton ("branch," "carry," "hurt," for heaven's sake?). I could go on, but why bother? Furthermore, the page is littered with misspellings, for example racoon, cumquat, attrium, and the aforementioned Avetsan. I don't blame the people who compiled the list, who are simply enthusiastic amateurs who love words and had no means of judging the validity of the etymologies they ran across, but it's depressing to think this is what people looking for information online will find and cite.
Here is a much more reliable list: only one word per language, but at least you can be pretty sure that word is correctly listed.
Addendum. A correspondent has brought to my attention
this silly site, which purports to list "some of Shakespeare's many coinages!" What they mean, of course, is "words first attested in Shakespeare," but that doesn't sound nearly as sexy. And some of them aren't even that; "accused," for instance, is centuries older:
1297 R. Glouc. 523 "Sir Hubert de Boru.. Acused was to the king of mani luther prise ['wrongful takings']."
Do these people really think Shakespeare
made up the word "alligator"?