French grammar question: use of articles in French version of the Constitution Act, 1867

I’m puzzled by the use of certain articles in the French version of the Constitution Act, 1867 of Canada: Loi constitutionnelle de 1867.

I had always been taught that provinces take a “de la” / “du” / "de l’ ", while municipalities take “de”, but that usage isn’t followed in the Loi constitutionnelle. “Nouvelle-Écosse” gets “de la” and “Nouveau-Brunswick” gets “du”, but Ontario and Quebec (province) both get “de” or "d’ ". See for examples the headers for arts. 112 to 116.

Is it because the provinces of Ontario and Quebec were new and it wasn’t yet established in 1867 how they would be treated grammatically, resulting from the splitting up of the “province du Canada”, which did get the “du”?

Another usage that puzzles me is the statutes dealing with Alberta, Saskatchewan and Newfoundland. The statutes dealing with Alberta and Saskatchewan are Loi sur l’Alberta and Loi sur la Saskatchewan, but Newfoundland gets Loi sur Terre-Neuve. Why not Loi sur la Terre-Neuve?

Je suis quelque peu perplexe.

“It’s like those French have a different word for everything.” - Steve Martin

I’m making a WAG here, but perhaps it is because Nouvelle-Ecosse and Nouveau-Brunswick took the names of existing countries in Europe (prefixed by “New”), so they shared the article that would be used for their European counterparts in long-established usage. Ontario and Québec, on the other hand, are original words from indigenous North American languages that did not have a previously existing European counterpart whose grammatical treatment they could borrow.

That may be it. Ontario and Québec are both masculine, which tends to be the default for words from other languages which don’t use grammatical gender. It may not have been established in 1867. (Except I have seen the construction “de Québec” in a statute from several decades later, in a sentence where today we would say “du Québec”.)

I’ve also noticed in the Loi constitutionnelle that there are some words with an accent aigu instead of an accent grave. The usage is consistent throughout, so I don’t think it’s an occasional typo: “siége” and “collége”, not “siège” or “collège”.

I wonder if that reflects an actual idiomatic pronunciation of Quebec French in 1867?

n.b. it has to do with shedding the English “province of Quebec”:

But that doesn’t explain why two of the original provinces got “du” / “de la”, while two other original provinces got “de” / "de l’ " when the provisions were using just the names of the province, not “province de”.

Sorry, typo. Meant to say:

From the Preamble:

And sections 6 and 7:

So it’s not the phrase “province de” which governs. Of the seven different provinces referred to in these provisions, five got “province du” or “de la”, including the provinces of Canada (the oldest name for the region now known as Quebec), and the province du Bas Canada, the intermediate name for Quebec.

It’s the two new provinces, anglo Ontario and franco Quebec, which get “province de / d’ “.

La province de Québec is perfectly idiomatic and is how it’s called to this day. *La province du Québec would definitely be incorrect. (According to @DPRK’s source it’s occasionally seen, but I don’t remember seeing it.) Of course, when not using the word province, it’s le Québec, so for example les lois du Québec (as opposed to les lois de Québec which would be the laws of Quebec City).

La province d’Ontario is a rare usage, but I can’t say for sure that it’d be incorrect today. Interestingly, unlike in the case of Quebec, la province de l’Ontario sounds just as correct to me. But both would be rare and today you’ll simply see l’Ontario. Since Ontario takes the article l’, I had to think for a bit to decide whether it actually is masculine, but it is: the expression le Nouvel Ontario used to refer to Northern Ontario but isn’t in common use today (and neither is le Nouveau Québec which used to refer to Northern Quebec).

I don’t know why Terre-Neuve doesn’t take the article, but it doesn’t. Ever. I will point out that there are other islands that don’t take the article, for example Svalbard, Maurice, or Madagascar, but it’s not consistent: l’Irlande, la Grande-Bretagne, l’Islande. Interestingly, je m’en vais à Svalbard, à Madagascar, à Terre-Neuve, but en Grande-Bretagne, en Islande.

As for siége or collége, I have no idea: to me those are clear spelling mistakes. Would they have been in 1867? Who knows, maybe accents weren’t completely fixed back then, in printed documents at least. Or it’s an OCR error.

So why la province de Québec but la province du Nouveau-Brunswick? I can’t say. If I had to make a guess, I’d say that the expression “province of Quebec” was created in 1763 with the meaning of “the province centred around and whose capital is the city of Québec”. It naturally became province de Québec in French, but the fact that it is named after the city was somewhat lost. However, you couldn’t have *province de Nouveau-Brunswick since there isn’t any city called Nouveau-Brunswick (well, not in New Brunswick anyway). Province d’Ontario, a later coinage, was then created on the model of province de Québec despite the fact that there isn’t any city called Ontario in Ontario. Is this a plausible explanation?

Thanks for responding, @Hypnagogic_Jerk!

Fair enough, but you do realise you’re shattering my image of French grammar as being all Cartesian and logical, as opposed to the helter-skelter that is English grammar? Customary differences??

The old “province de Quebec” only existed for 28 years, before being replaced by the “province du Bas-Canada” in 1791 (at least, as that phrase is used in the Loi constitutionnelle), so I wonder how much continuing currency that usage had over 70 years later, in 1867?

Okay, islands are grammatically different, except the statute in question is referring to the province of Newfoundland, not the island of Newfoundland. The province of Newfoundland is made up of the island of Newfoundland and the mainland of Labrador. So we’ve got a third grammatical treatment of a province, based on the fact that most of the province’s population lives on an island?

I don’t think it’s an OCR error, because it’s too consistent. But from personal experience, I found the accent for Quebec City to have a very strong “e” pronounced as “ay” (sorry, i can’t do the fancy pronunciation code). I wonder if it was a regional accent? I’m reading the 1867 statute version, and there were other words with an e grave, so it wasn’t that the type-set was lacking.

That may be it, combined with Schnitte’s suggestion about lack of grammatical gender for Quebec as a region?

But it still leaves the other examples in my OP, which don’t turn on the usage of “province de”. The marginal notes for ss. 113 to 115 are:

It seems like a general usage in the Loi that Ontario and Quebec take “de” / "d’ ", not one that turns on the term “province de” / “province du / de la”.

All very puzzling, but I appreciate your comments.

I’ve looked a bit at the Constitution Act on the official ministry of Justice’s website, as linked by @Northern_Piper, and there are a few oddities there. For example, article 109 says “Toutes les terres, mines, minéraux et réserves royales […] appartiendront aux différentes provinces d’Ontario, Québec, la Nouvelle-Écosse et le Nouveau-Brunswick […]” which is just incorrect. What I would say is “appartiendront aux différentes provinces de l’Ontario, du Québec, de la Nouvelle-Écosse et du Nouveau-Brunswick” which I guess is therefore one situation where I would consider “province […] du Québec” to be correct. As for “Actif d’Ontario et Québec”, I would instead go with “Actif de l’Ontario et du Québec”. Then again, legal documents are sometimes written in a formal style that’s different from the one we’re used to. As for article 105, it refers to the “Royaume-Uni de la Grande-Bretagne et d’Irlande”, where to me this “lais definitely superfluous. (It also calls the pound sterling the “louis”, but that’s not the only interesting thing about currency names in this document; see next paragraph.)

Now looking at articles 112 to 116, what strikes me isn’t the grammar but the fact that they officially refer to the Canadian currency unit as the “piastre”. (The English version calls it “Dollar”, with a capital D.) Piastre is a synonym of dollar that’s been in use since it referred to the Spanish dollar, it’s (informally) used in Quebec and Canadian dialects of French to this day, and I’ve also seen it in the comic Lucky Luke (written by Belgian Morris and Frenchman René Goscinny) and in an earlier edition of the French version of the Haxby and Willey guide to Canadian coins, where, in the historical section, they refer to the “dollar-cent-mill” system as “piastre-centin-millin”. (In European dialects of French, the hundredth part of the base monetary unit is usually called un centime, while in modern Canadian French it’s formally un cent, which can informally become feminine and even be written une cenne.) However, to me piastre is mostly an informal term, and I wouldn’t have thought it would have been the one formally used by the drafters of the Constitution Act.

Grammar is going to treat Terre-Neuve similarly regardless of whether it refers to the island or to the province. Ireland is another word that refers both to an island and to a country that aren’t coterminous, and it doesn’t take a different preposition depending on whether we mean the Republic or the whole landmass where Dublin and Belfast are located.

However, I do feel that there is something strange about Terre-Neuve. Suppose someone called Roger buys an island and names it after him. (In fact, it doesn’t even need to be an island, it can be any piece of land that becomes officially named after him.) If he calls it Rogerus, then I would naturally say in French je m’en vais à Rogerus. However, if he calls it Rogeria, I would say la Rogérie and je m’en vais en Rogérie. This should also convince you that, even though exceptions exist, grammar is actually quite regular and predictable. Terre-Neuve looks like it should be in the second category (like l’Irlande or la Grande-Bretagne), but actually acts in a way that places it in the first one (like Svalbard or Bornéo). I’m not sure why that is, but it could be that Terre-Neuve lost its meaning of la terre neuve and started being seen as a single atom, so to speak, in the same way that in English Newfoundland isn’t pronounced the way “new found land” would be.

Also, le Rogeristan and je m’en vais au Rogeristan. (As well as au Turkménistan, but en Afghanistan.) Now I’m thinking that there probably exist linguistics papers about the grammar of place names in French that could shed light on this phenomenon. I may try to look for one sometime later.

The closed e /e/ and open e /ɛ/ are two different phonemes in French, but in some dialects they tend to merge into a single sound that’s somewhere in between. However, Quebec or more generally Canadian dialects tend to preserve the distinction moreso than others. Also, it shouldn’t affect the spelling, especially not in formal documents. Where did you see this, by the way? Articles 68 and 72 on the official ministry of Justice’s website, for one, spell “siège” and “collège” the way I’d expect to see them.

Fascinating, thanks!

I’ve been using the original French version, printed in the federal statutes in 1867:

The version used in the DOJ website has been tidied up a bit, correcting some other typos and fixing the accents aigu to « siège », and « collège ” ».

In addition to « piastre », the Act also uses « centins », in s. 118, the “eighty cents a head” subsidy to the provinces.

And in ss. 146 and 147, there is yet another differing usage of the article, in regards to islands: « provinces de Terreneuve et l’Ile du Prince Edouard ». Two islands, but one with an article and one without.

All very interesting - thanks for your comments, Hypnagogic Jerk. I"m just coming to the conclusion that French is not as orderly as I had assumed it was.

“The French don’t care what they do actually, as long as they pronounce it properly.” - Prof. Henry Higgins

Okay, I’ve searched for the words “siège” and “collège” in the Trésor de la langue française informatisé, which provides historical and etymological information. Here are the relevant lines:

  • For siège: Prononc. et Orth.: [sjε:ʒ]. Ac. 1694, 1718: siege; 1740-1835: siége; dep. 1878: siège.
  • For collège: Prononc. et Orth. : [kɔlε:ʒ]. Ds Ac. 1694-1932. Écrit collége ds Ac. 1762-1835 : ,on prononce colége``. Pour cette graph. cf. aussi Nod. 1844, Besch. 1845 et Lar. 19e ; écrit collège pour l’Ac. à partir de Ac. 1878 suivie par le reste des dict. gén. Cette graph. est déjà celle de Fér. 1768, Fér. Crit. t. 1 1787, Land. 1834 (qui transcrit cependant [e] fermé) et Gattel 1841.

So the word “siège” was spelled siege in the first two editions of the Dictionnaire de l’Académie française of 1694 and 1718, siége in the next four published in 1740, 1762, 1798 and 1835, and siège starting with the seventh edition of 1878. As for “collège”, it was spelled collége in the 1762 to 1835 editions and collège afterwards. However, the spelling of this word was more debated since the modern spelling was already used in other dictionaries since the 18th century. However, the pronounciation used a closed [e] (which would suggest a ‘é’ spelling) instead of an open [ε] until at least the 19th century, even according to some dictionaries otherwise recommending the “collège” spelling.

That’s interesting, thanks. So it looks like the accents changed on “siege” right around the time of the 1867 Act, and therefore it was an acceptable usage, while “college” was debated?

I guess the DOJ just decided to update to the modern accents.

Thanks for digging that up. Fascinating.

It’s possible that the spelling of siège was also debated; the Trésor also shows uses from the 19th century where it appears to be written with a grave accent, although maybe they’ve modernized the spelling. The Dictionnaire de l’Académie française only adopted the modern spelling (for both words) in its 1878 edition, so if the drafters of the 1867 Constitution Act used as reference its 1835 version, that would explain their choice of spelling. But I don’t know if that would also have been the more common spelling among educated people in Canada at the time, nor what dictionaries were in common use. The Dictionnaire de l’Académie française is known as fairly conservative, but I also don’t know if it was the case back then.

There are other interesting word choices in this document, such as the word “bill” being used with the same meaning as the English word. But now that I search for it, it’s actually found in the Trésor with a reference from the 1835 Dictionnaire de l’Académie française. Modern politicians in Canada would say projet de loi, however.

Yes, there is a bit of a discussion about the wording of the 1867 Act. It has been criticised for the anglicisms, such as the use of “bill” instead of “projet de loi” and “acte” instead of “loi”, the latter as a faux ami. As part of the patriation project, the federal government directed that the unilingual English parts of the Constitution should be translated into French and then enacted. The versions produced by the drafting committee are also on the DOJ website, including a new version fo the 1867 Act:

However, that version, and the translation project in general, have been critised by Professor Choquette at Queen’s, as trying to re-write the historical version, thereby deny the legal and linguistic heritage of French-Canadians from the Confederation period, and anachronistically reading in subsequent constitutional concepts that were not in existence in 1867, such as “dominion status”.

He also makes the interesting point that although the 1867 version was drafted by Eugène-Philippe Dorion, there is some indication that George-Étienne Cartier also had a hand in it, which increases the significance of the Dorion version as reflecting the views of a key Father of Confederation.

Choquette’s paper is available here:

So if I understand correctly, the original French version printed in the federal statutes in 1867 does not have official status as law. What about this translation that was prepared following the 1982 patriation, has it been officially adopted? And is the version on the Departement of Justice’s website that you linked to in the OP the same as this modern translation or yet another version?

None of them have official status, because the British Parliament only enacted the BNA Act (as it was originally named) in English.

Dorion then produced the French version, which was published in the French volume of the 1867 statutes. It’s been cited ever since by the courts, so it seems to have some customary law status.

The DOJ version on the main page is the Dorion version, with some typos cleaned up and the accents updated. Howver, that version keeps the original usage of articles which started this discussion.

The new version produced by the French drafting committee is just a draft, and has no official status. Parliament alone cannot bring it into force. To have constitutional status, it would need to be enacted by the amending formula set out in the Constitution Act, 1982. That hasn’t happened, probably because it would likely need unanimous consent of all the provinces. My guess (just a guess, haven’t seen this set out anywhere) is that Quebec would have difficulty agreeing to do so, because that might be seen as ratification of patriation. For whatever reason, those new drafts have been sitting there for almost thirty years, not enacted.