What's affected by 'Use American dialect'?

Grepping through my Inform 7 tree, I see that ‘Use American dialect’ suppresses the inclusion of “and” in numbers such as “one hundred and seventeen”, but I don’t it used anywhere else. The manual says that also causes American spellings to be used, but I can’t find any examples, in the standard rules or anywhere else, where it actually has this effect. Is there one I’m missing?

The reason I ask is that I’m writing a game set in the Yukon Territory, and as such I’d like it use Canadian English. Canadians use British spellings in some cases (“colour”, “centre”) and American spellings in others (“tire”, “realize”). I’m pretty sure that for the standard rules the default dialect works fine (judging by the Strathy corpus, most Canadians do say “and” in the pronunciation of large numbers), but I want to make sure I’m not overlooking anything.

On a search through the Standard Rules, there’s “recognis/ze” in the “That’s not a verb I recognize” response and “apologis/ze” in the standard response to “sorry” (“Oh, don’t apologize”). Those are the only two I found, though I don’t guarantee that there aren’t any that pop out of verb conjugations in Preform.

Grepping the Standard Rules shows two places: “That’s not a verb I recogni[sz]e.” and “Oh, don’t apologi[sz]e.” It may also be used in some other extensions though I can’t think of any specific examples.

EDIT: What Matt said, though I don’t believe this affects automatic verb conjugations.

Thanks. Since I missed both of those, I must be grepping for the wrong thing; I’ll check again. Canadian English uses a ‘z’ in both those cases so now I guess I’m into manually fixing up rules since neither built-in dialect quite suffices. Easy enough. I’ll fix those two and any more that I find.

Look for “American.”

Pretty sure there’s nothing in there that would be Canada-but-not-US-or-vice-versa in the text.

There’s also a check when printing large numbers in words, whether to say “two hundred one” or “two hundred and one”. The “and” is considered UK dialect. (I dunno how well that matches up with people’s usage, I could go either way with it.)

Not strictly true. Canadians can use either ‘realise’ or ‘realize’, and the ‘ize’ endings have gained currency only recently. All my English teachers in the '70s and '80s in Toronto taught me to spell it ‘ise’ and I’m taking that to the grave. Can’t speak for the Yukon specifically, but if you’re a Canadian you have your choice of ‘ise’ or ‘ize’ and neither should be considered ‘incorrect’.

Well, the two words at issue here are “regcogni[sz]e” and “apologi[sz]e”. In the Strathy Corpus I see 3328 uses of “recognize” versus 93 of “recognise”, and 525 of “apologize” versus 9 of “apologise”. That’s a pretty strong skew toward -ize. On the matter of using “and” in numbers it’s hard to get an exact count out of the provided serach facilities, but I see a lot more with than without, and all my Canadian friends insist that the “and” belongs there. So whether or not other choices are “incorrect” (and I don’t think any of them should be considered such no matter where in the world you live), these choices are clearly the most common and are what I’m going to use.

I’ve achieved the behavior I want out of Inform by using the default dialect to get the “and”, and then adding the following two rules to change the spellings to -ize:

The parser error internal rule response (N) is "That's not a verb I recognize.".
The block saying sorry rule response (A) is "Oh, don't apologize.".

I wasn’t trying to push you toward using ‘ise’ - just informing you (since you say you don’t live here) that ‘ise’ is considered ‘a’ Canadian usage. But hey, I’m sure an internet search knows better.

BTW, my understanding is that the situation in Britain is identical to that in Canada - both ‘ize’ and ‘ise’ are considered acceptable and correct; it’s just that in Britain, ‘ise’ is still the more common variant (although - I’ve heard that’s changing there, too). Only in America is one of those two considered ‘incorrect’. Sometimes people try to reverse this situation, as if only ‘ise’ is correct in Commonwealth countries (esp. Britain, Australia, NZ), but that doesn’t really fly because both forms are equally ancient - neither has ‘old world’ priority — and both tend to see actual use.

I have no opinion as to ‘and’ in that context. I would probably insert it, myself. Sounds awkward to my ear, otherwise, but I wouldn’t call it un-Canadian.

This is what I do for adaptive cross-dialectical spellings. This system can easily be hacked to make special exceptions for Canadian spelling (although I just go with British b/c there are very few differences btwn British and the Canadian I was taught as a child - basically ‘tire’, ‘maneuver’, ‘jail’ I was taught to spell American-style instead of ‘tyre’, ‘manoeuvre’, and ‘gaol’, but pretty much everything else, British-style)…

To say ou:
     say "[if American dialect option is active]o[else]ou[end if]";

To say iz:
     say "[if American dialect option is active]iz[else]is[end if]";

To say ll:
     say "[if American dialect option is active]l[else]ll[end if]";

To say nc:
     say "[if American dialect option is active]ns[else]nc[end if]";

To say ys:
     say "[if American dialect option is active]yz[else]ys[end if]";

To say re:
     say "[if American dialect option is active]er[else]re[end if]";

To say ered:
     say "[if American dialect option is active]ered[else]red[end if]";

To say ering:
     say "[if American dialect option is active]ering[else]ring[end if]";

To say ogue:
     say "[if American dialect option is active]og[else]ogue[end if]";

The parser error internal rule response (N) is "That's not a verb I recogn[iz]e.".
The block saying sorry rule response (A) is "Oh, don't apolog[iz]e.".

This should adapt to the ‘Use American dialect’ option. You can use substitutions to adaptively get more differential spellings like ‘col[ou]r’, ‘lice[nc]e’, ‘paral[ys]ing’, ‘cent[re]’, ‘cent[ered]’, ‘cent[ering]’, ‘catal[ogue]’, and ‘labe[ll]ed’.

BTW be aware that as of v6L38 Inform 7’s adaptive verbs do not seem to be dialect-aware, at least judging by the handling of the verb ‘to pedal’, which by my understanding should be conjugated ‘pedalled’ in Canadian/British usage and ‘pedaled’ in American usage, whereas Inform gives us ‘pedalled’ either way — better watch those conjugations if you’re using adaptive text and going for authenticity b/c I think it’s a lot less likely to see a Canadian write American-style ‘pedaled’ than ‘apologized’.

[ EDIT P.S. Just remembered: The word ‘practice’ is a weird singular exception. As a verb only, Americans spell it ‘practice’ whereas Canadians spell it ‘practise’. (We both spell it ‘practice’ as a noun.) It’s the only word that I could find where the Americans — unlike with offence/defence/licence — favour ‘ce’ over ‘se’ (I’d be curious if there are more!) so rather than generalise it, I just handle that word individually. ]

As far as “pedaled” vs “pedalled”: I was taught that American usage is to double the final consonant[1] regardless of the length of the word when a suffix is applied (pedalled, travelled, worshipped, programmed) whereas the British would use a single consonant if the word were more than one syllable (pedaled, traveled, worshiped, programed). As an American the former spellings look more correct to me while the latter seem a bit strange. It sounds like the double-consonant forms are more prevalent in Canada as well then?

[1] though -cc- sometimes becomes -ck-, picnicking instead of *picniccing

Sorry, I didn’t mean to give offense, and anyway, it sounds like we’re in violent agreement.

(Now you’ve done it… I had to pause for several seconds on the word “offense” in order to remember my own native dialect).

Nice. I considered something like this but decided it was overkill: I’m just writing a game, not an extension, so once I have the built-in messages behaving the way I like I can just hard-code everything else. If I needed my prose to be reusable I’d take your approach, though.

The only word I can think of which Americans spell with -ice as a noun and -ise as a verb is “advi[cs]e”. Some words where Americans keep the -ice spelling in both noun and verb form are: apprentice, invoice, notice, prejudice, price, sacrifice, service, voice; AFAIK all eight of these spellings are universal across dialects.

This Oxford Dictonaries blog entry suggests that “device/devise” is another such pair, though the meaning of the noun “device” has drifted pretty far away from that of the verb “devise.”

Maybe also related is the noun “prophecy” and the verb “prophesy.”

I was taught essentially the opposite. And even if you check today’s grammar guides and usage manuals, what they often say is, everybody doubles consonants like ‘T’ and ‘L’ on words where the final syllable is stressed (e.g. ‘gelled’, ‘corralled’) but Americans don’t double them when the final syllable is unstressed, which explains why [EDIT: many] Americans write ‘canceled’ and ‘traveler’ whereas I would write ‘cancelled’ and ‘traveller’.

I have never seen ‘traveler’ in non-American usage — only ‘traveller’, so if somebody has said that anybody but Americans tend to write it with one L I’d find that odd and not matching my own experience.

As far as I know, those spellings don’t differ across ‘dialects’ (which is a misnomer), though, so when Americans spell the nouns ‘device’ and ‘advice’, but spell the verbs ‘devise’ and ‘advise’, it’s because every other English speaker spells those nouns with ‘C’ and those verbs with ‘S’, too. Therefore, adaptive algorithms would not be necessary.

‘Practice’ as a verb is the only instance I could find where Americans opt for the ‘c’ where everybody else opts for an ‘s’. Judging from the rest of their spelling conventions (dropping the verb/noun distinction for ‘license’, ‘offense’, etc), I’d have expected them to simplify toward the ‘s’ instead, which is why I find it counterintuitive and am interested in a similar example. I don’t think there are any. I think ‘practice’ stands alone.

Interesting. I’ve almost always seen the doubled consonants in local usage as well, but I’d thought it was an American influence that was spreading elsewhere (hence the ubiquity of doubled consonants in computer-related terms: I’ve almost never seen *programer, *formated, *inputing, *outputing). But I checked my style guide and it agrees with yours: the doubling is more common in British English.

Looking at the list of examples they give, my idiolect does seem to be somewhat inconsistent. I would write worshipped, targetted, modelled, but focused instead of *focussed. So perhaps fricatives are an exception? I’m trying to think of words ending in an unstressed vowel followed by a single -f or -z now and coming up blank…

There are indeed plenty of exceptions. Or perhaps more accurately, the ‘don’t double unstressed consonants’ rule sometimes used in America only applies, as I understand it, to some consonants and not most. I know it applies to T and L - but I don’t think it applies to M or N - AFAIK nobody on the planet makes a habit of writing ‘programer’ or ‘shadowbaning’. Not as sure about P or B, but I would think not. Consonant doubling is wildly inconsistent in all forms of English, but I think the only consonants for which we need to consider doubling differences are T and L; I might have missed a trick or two, though.

You appear to be correct for all practical purposes. Grepping through some wordlists led me to one other deeply obscure example. The word “grice”, or “grise”, is a synonym for “piglet”. It appears with both spelling in my British wordlist but only with a “c” in my American wordlist. Webster’s 1913 suggests that “grise” is used in Scotland and “grice” everywhere else, but the word is so rare that it’s not in the OED or any of the BYU corpora except as a surname or a misspelling of another word.

Very interesting. I know ‘grise’ as a French word meaning ‘grey’ (grey/gray, there’s another singular difference), but I have never heard of the English ‘piglet’ form. It wouldn’t form the basis for a generalised rule, anyway, so I guess I’ll keep handling ‘practice’ as an isolated case. (It doesn’t really bother me to do so — I just have formed this habit of trying to generalise solutions wherever possible.)