Linguee: translation, but also shows you context (lots of places where the word or phrase is used) as an aid to learning
Netspeak: a nifty little tool that uses web search to tell you if phrase X or Y is more popularly used
Google Labs presents the Books Ngram Viewer: chart/compare the usage of words/phrases in books over time
The Great Language Game: I can really only narrow it down to language group, and then I'm flipping a coin
Using may have vs. might have: "The alarm system may have prevented further damage" means that the alarm system did go off, people responded to it and we think things would have been worse otherwise. "The alarm system might have prevented further damage" means either it didn't go off at all, or it went off, but in a way that was ineffective, but whatever the case, it didn't work and it's too late now. H/T: Comic Strip of the Day
Getting your locale set right is important, but if you don't, a dedicated post office may compensate for you.
The Secret Language Code: it turns out that statistical analysis of a writer's use of pronouns reveals that they are correlated to lots of demographics about the writer, e.g. gender, age, social class, etc.
Sign language that African Americans use is different from that of whites (even at Gallaudet)
Why is [state X] so [Y]? as revealed by Google searches
Stephen Fry on Language (kinetic typography): big reveal at the end
The McGurk Effect: what we see can override what we hear (even when we know about the effect)
Lexicalist ("a demographic dictionary of modern American English"): breaks down words (e.g. found in Twitter) by age, gender, and geography
For all but one inflection, there are two ways to express the negative contraction form of "to be", e.g.: "you aren't" and "you're not", "he isn't" and "he's not", etc. "I" is the exception, which has only one form: "I'm not". Now: When the negative contraction of "to be" is placed at the end of a sentence, as a question-confirmation, it has to use the "n't" form, e.g.: "You're planning to cheat, aren't you?", "He's very smart, isn't he?". The trouble is that there's no "n't" form for the "I" case. Since we can't say "amn't I" we end up using "aren't I", e.g., "I'm going to fail, aren't I?" — which is not grammatically correct. (I'm trying to break this habit myself.) Captain Von Trapp sets the example for us. [Who thinks about this stuff?]
Negative questions in English are generally a bad idea. If you say, "You don't have any money?" and I reply, "No," am I disagreeing with you because I do in fact have money, or am I asserting I have no money and thus agreeing with you? Why ask a question that's likely to require an immediate follow-up question? Apparently German has a way to handle this: In addition to ja and nein you can answer doch — which always carries the meaning that you are contradicting what the speaker said or asked.
The Unrecognized Death of Speech Recognition
MLA interactive map of 'speakers per language in the USA'
Very insightful article on why Google offered free GOOG-411, the new Android speech-search app, etc.: It's to build a large phoneme database to improve speech-to-text which will then allow them to index video (YouTube).
Sometimes Google doesn't help when all you have to go on is a phonetic exchange. From That Thing You Do:
Guy: "If Jimmy's a genius then I'm [oo tahnt]."Try Googling "Oo Tant" or "Ou Tante" etc. and you'll get lots of pages talking about TTYD, but nothing else. Turns out the reference is to U Thant, the 3rd Secretary-General of the United Nations, 1961-1971, from Burma (now Myanmar). "'U' is an honorific in Burmese, roughly equal to 'Mister'." — Wikipedia
TBP: "Who's [oo tahnt]?"
Guy: "He's the sec— Forget it."
My contact information, brief autobiography (such as it is), and family tree. back to Nordist.net