I tend to ignore Bryan Garner’s Garner’s Modern American Usage, a beautifully produced and widely used and respected book with whose prejudices and general approach I utterly disagree, but recently I wanted to see what he had to say about something and found myself instead looking at his long list of “Denizen Labels” — what are usually called demonyms. Now, I’m very fond of demonyms myself, though I’ve only devoted one post to them (and there are lots of great examples in that thread); my copies of Daniel Santano y León’s Diccionario de gentilicios y topónimos and A.M. Babkin and E.A. Levashov’s Словарь названий жителей СССР [Dictionary of names of inhabitants of the USSR] are treasured possessions. I therefore dropped what I was doing and pored over the list, quickly concluding that it showed the same combination of random choices, poor decisions, and sloppy thinking that irritates me so much about the rest of the book.
An example of random choices: why does he include Dublin (Dubliner) but not Cork (Corkonian)? It’s certainly not because Cork isn’t the capital; he has lots of fairly insignificant places, like Metz (Messin), Saint-Cloud (Clodoaldien), and Trois-Rivières (Trifluvien) — he even has Dundee (Dundonian), to which Cork (Corkonian) would make a nice companion. Poor decisions: for Budapest he gives “Budapestiek,” which is a capitalized version of the Hungarian plural for budapesti ‘inhabitant of Budapest.’ If you’re too ignorant of a language to distinguish singular from plural, you shouldn’t be trying to provide a demonym from that language. And that brings us to the third issue, sloppy thinking: who are these terms intended for, and what use does he envisage? He seems to have extrapolated from French demonyms, which he clearly loves and which in fact are often used in English, the idea that one should use native terms wherever possible, but he’s inconsistent about this: he gives (besides that stupid Hungarian plural) Istanbullu for Istanbul, but for Helsinki he gives Helsinkian, not helsinkiläinen (or, as he would write it, Helsinkiläinen). No English-speaker is going to use “Istanbullu” unless they live in Turkey and are immersed in Turkish culture; he’s just showing off. The Wikipedia entry linked above, though not complete (they don’t have Helsinki, for example), is more sensible; for Istanbul they give Istanbulite, for example. And again he’s inconsistent: for Dijon he gives Dijonese, not Dijonnais, the French term. For Shanghai he gives Shanghailander, which sounds archaic to me; Wikipedia gives Shanghainese, which is what I would say myself. In short, his list, like his book, is an impressive-looking but antiquated and incoherent piece of work.
Update (March 2016). Having gotten a review copy of the new Fourth Edition of this book, I naturally turned to the “Denizen Labels” and found that Garner has added Cork (Corkonian), changed “Budapestiek” to “Budapesti” (which at least is singular), added Istanbulite to Istanbullu, and added Dijonnais to Dijonese. Hey, maybe he read this post!