Nobody in Middle English could decide how to standardize the spelling of crocodile, which gave us outlandish variants including cokedrille, kokedrille, cocodrill, cokadrille, and cokedril, until our current spelling won out. All this comes from French cocodril, from Medieval Latin cocodrillus, from regular Latin crocodilus, and finally from Greek krokodilos, which is closer to the modern word than most of its descendants, because crocodile as we know it was kind of conservatively refashioned to look like its Latin and Greek ancestors. This is where the word was created; it makes sense that the Greeks got to name the crocodile first because during the Hellenistic expansions they encountered them in Egypt. Anyway, the animal was named by Herodotus, who combined the words kroke, or "pebbles", and drilos, or "worm". Sadly, further etymologies of both words are obscure and untraceable. However, we learned something interesting nonetheless!
The word assassin, I assumed, would not be Indo-European, based off its unusual formation. This conclusion was correct; it is from Arabic, by the normal route through Italian then French (in this case by assassino and the homonym assassin, respectively). But what is curious here is the semantic change: in Arabic, the word was pronounced something like hashshashin and literally meant “hashish addicts”, hashish, of course, being a form of marijuana. The connection exists in an 11th century Ismaili cult led by a man named Hasan Ben Sabah, a group which would send out its highly trained initiates to kill opponents- only after rituals where they had to intoxicate themselves with cannabis. Hashshashin, of course, is connected to the word hash (slang for “weed”, not to be confused its homophones, which have different etymologies), a shortening of hashish (still in use, though even less common now). All of this comes from Arabic hasis, or “dried herb”, from a Proto-Semitic source. In the transition from hashshashin to assassin, you may have noticed all the h’s were dropped. It is relatively common for an sh to become just an s, and a word starting with h and a vowel to lose the h; see my post on humble pie about the opposite effect. But, anyway, I’ll never look at a murder mystery show the same way again.
There are some people out there who will tell you that the word barbecue comes from French barbe a queue, or "beard to tail". Don't listen to those people. The noun is from the 1690s, but the verb is from the 1660s, and it came from Spanish barbacoa, not French. This is a remnant of Spanish occupation of the Caribbean, from either southern Florida or one of the greater Antilles, where natives used a word like barabicu or barbakoa (accounts vary) to describe their practice of roasting animals on a grid of wooden sticks. Metynomically, the word originally meant "wooden framework" in general. This is from one of the Arawakan tongues, from a theorized Proto-Macro-Arawakan source language. With no writings, it's hard to philologize such languages. Anyway, after barbecue was passed from Spanish American colonists to English American colonists, it obtained several different spellings and abbreviations, like bar-b-q, BBQ, barbeque, 'que, 'cue, and barb-b-que. The q for c substitution is partly because of that folk etymology I already discussed and partly because it just sounds like it should be spelled that way.
Cattywampus is a curious colloquial American word doubling as meaning "diagonal" and "askew". The etymology is a mongrel. The "diagonal" bit may have been influenced by catty-corner, and wampus could have been southern slang for "flail", but then there's a fictional creature called the wampus cat and another called a catamount. Then again, wampus might derive from the Scots word wampish, meaning "twist" and catty may have been a formation from cater, not the one you think but an obsolete term for "diagonal" also present in catty-corner. It has also been suggested that the whole thing was made up to sound like a funny yet classical word, with catty possibly harkening back to Greek kata and the second bit just sounding funny. Just as nobody can agree on the etymology, no one seems to concur on the spelling; in addition to cattywampus, there are existing forms such as catawampus, kittywampus, catiwampus, catawampous, caddywhompas, and much, much more. This is because cattywampus seems to be a two hundred year old slang word passed down verbally, and no one can agree on standardization. Usage began in the early nineteenth century, peaked in the mid twentieth century, and has been increasing in the twenty-first century so far.
In Middle English, the word welcome went through many alterations, taking unrecognizable forms such as wolcume, wulcume, wilcume, and wilcuma. All, however, carried the same definition and all trace to the Old English term wilcuma, a noun literally meaning "one who arrives at the pleasure of another". This is a combination of willa, or "pleasure", and cuman, or "to come" (yes, the word come comes from this, through a couple more similar spellings in Middle English as well). Willa, through Proto-Germanic wiljana, goes back to Proto-Indo-European welh, or "to wish". Another interesting part of willa is that while it is related to our word well, it's closer so to will, and the modern modification that looks like well in welcome is actually an instance of folk etymology; people thought that that was supposed to be the correct spelling, and changed it to be so. Instead of hoping someone comes well, the word wills them to come, but that's largely forgotten today. Meanwhile, cuman, through Proto-Germanic kwemana, traces to PIE gwem, which meant "to step". You can see the "traveling" semantic connection. So, welcome etymologically means "wish step", because it bids people to come forward. So how did this get adopted in you're welcome? When you reply with that nicety, instead of inviting them to come, the meaning shifted to inviting them to ask for such a favor again. Today we don't even consider it. Well, we'll come around.
Adam Aleksic, a freshman studying linguistics at Harvard University, has been described as the internet's sixth most famous etymologist. He also has disturbing interests in words, vexillology, geography, board games, limericks, and law.