Language Log

The first English-language slang dictionary:
A New Dictionary of the Terms Ancient and Modern of the Canting Crew, in Its Several Tribes, of Gypsies, Beggers, Thieves, Cheats, &c, With an Addition of some Proverbs, Phrases, Figurative Speeches, &c Useful for all sorts of People (especially Foreigners) to secure their Money and preserve their Lives; besides very Diverting and Entertaining, being wholly New.

Most words have different senses and meanings depending upon the context in which they are used, hence the need for multiple definitions in dictionaries.  Philip Taylor has come across this article about:

8 Words That Are Only Used in One Weirdly Specific Context
Think about it: have you ever heard someone say they had “extenuating errands”?
By Sam Hindman, Mental Floss (11/23/25)

Yesterday afternoon, Andrej Karpathy posted llm-council on github:

The "words per day" thing has mostly gone quiet since it erupted in the media nearly 20 years ago — but here's a sensible replay:

I have observed the author working on this 749 page volume for many years, so it is with great rejoicing that it is available in time to send to friends, colleagues, and students as a Yuletide gift:

South Coblin, Common Shē and Common Hakka-Shē: A Comparative Study
Language and Linguistics Monograph Series 68

Institute of Linguistics, Academia Sinica (Taipei:  November, 2025)

Introduction

The present work is divided into two parts. Part I is devoted to the reconstruction of the phonology of Common Shē, the ancestral form of the closely related Sinitic dialects spoken by the Shē ethnic minority of China. The approach applied is the classical comparative method, in which modern data from seventeen modern dialects are subjected to comparative reconstructive analysis. Data from additional Shē varieties are also adduced as needed. The end product of these procedures is a hypothetical phonological system, which for the sake of brevity we call Common Shē, though this term should more precisely encompass not only phonology but also syntax and lexicon.

In our discussion of part III of Elle Cordova's Grammarian Saga, "Grammarian vs Errorist showdown at the secret L'error", there was some back-and-forth about whether or not error could stand it for the nucleus of lair.

AntC wrote "I can't smudge 'error; into a single syllable", and I responded that "Elle Cordova does it by brute force — I don't think 'air' is her normal pronunciation of 'error'".

But looking into pronunciations of error on the internet, I've found that her "l'error" ~ "lair" pun is, in fact, pretty plausible.

In an email, Coby Lubliner observed that many English words have more than one standard pronunciation. He asked whether there has been any academic discussion of this general issue, and noted that he isn't familiar with any other language with as much variety in its standard pronunciations.

I, for one, don't think it's the least bit funny.

Uyghur pronounciation

The way Uyghurs speak Mandarin is now a joke
For many it’s not funny, given the political heat around language choices

Economist (Nov 13th 2025)

The article begins with a viral joke, which Economist doesn't bother to explain (I will, though, at the end of the first paragraph):

Scroll through posts about Xinjiang on Chinese social media and an odd phrase soon appears: “Apple U”. It is a pun that mimics how some Uyghurs, the largest ethnic minority in Xinjiang, a region in China’s far north-west, pronounce “Hey, friend” in Mandarin. This meme is part of a growing trend online for using nang yan wen, or “naan Mandarin”—a way of writing and talking that wags have named after Xinjiang’s staple flatbread. Videos tagged with the term have amassed more than 1.7bn [VHM: !!!] views on Douyin, TikTok’s sister app in China, since the start of the year.

Fritz Newmeyer suggested that I post this.

The Linguistic Society of America will be offering a quasi-hybrid registration option for the 2026 Annual Meeting, to be held January 8-11, 2026.

For those of you who cannot travel to New Orleans in person, attending remotely through the quasi-hybrid registration option will give you virtual access to all plenary talks, plus access to four concurrent session rooms where a continuous schedule of sessions selected to reflect the full array of disciplinary subfields will be offered across all four days of the meeting.

As a virtual attendee, you will be able to see the speakers, ask questions, and hear questions from the live audience. This year’s virtual option is a pilot test for a cost-effective approach to hybrid meetings. If it is successful, we will expand the number of hybrid rooms in coming years. Please tell your international colleagues or anyone who is concerned about traveling for any reason about the virtual registration option and encourage them to join us!

You can filter the 2026 LSA Annual Meeting Program for the Track "Hybrid" and see the 45 sessions that virtual registrants will be able to attend, which includes all plenaries plus four continuously scheduled concurrent session rooms.

The links to access the sessions will be in the meeting app, which everyone who registers for the meeting will be able to download.

The genius logic of the NATO phonetic alphabet (title of the YouTube video)

In the comments on "Final prepositions again", AntC alerts us to Elle Cordova's latest, part III in the Grammarian Saga: "Grammarian vs Errorist showdown at the secret L'error".

"Dictionary.com’s 2025 Word of the Year Is…"

Each year, Dictionary.com’s Word of the Year and short-listed nominees capture pivotal moments in language and culture. These words serve as a linguistic time capsule, reflecting social trends and global events that defined the year. The Word of the Year isn’t just about popular usage; it reveals the stories we tell about ourselves and how we’ve changed over the year. And for these reasons, Dictionary.com’s 2025 Word of the Year is 67.

Macquarie Dictionary's WOTY shortlist also included six-seven; Sam Altman is apparently planning to name his next AI model GPT-6-7; and a news search will give you plenty of other relevant stories, from basketball scores  to "6-7 in the Bible".

Since Victor recently spent 1100 words on various people's "best approximations of how they think they are saying 'Canton'", "expressed in common spelling (not a phonetic alphabet)", and has resisted requests to provide audio, I thought I'd provide some examples of how a Canton resident pronounces the city's name. As I've explained many times, I don't think that IPA transcriptions are an effective way of representing how people actually talk, and this case will continue to support that view. Instead, a good place to start is a sample of audio clips along with graphical representations of waveforms, spectrograms and other kinds of acoustic analysis — and there are several possible directions to go from there.