The beginning of the end: The Old French conjunction ains in the sixteenth and seventeenth centuries
Juhani Härmä, Department of Modern Languages, French Philology, University of Helsinki
The present study discusses the semantics and pragmatics of the French conjunction  ains and its connections with what is considered as its modern French equivalent, mais. In addition, the study makes some comparisons with certain other languages (mainly languages which are known to have equivalents of the two conjunctions mentioned (see the beginning of section 3). Beside discussing the functions of ains, the second question I wish to address here is the somewhat intriguing disappearance of this particular conjunction. Ains appears in Early Old French and survives through Middle French and Renaissance French until the beginning of the era of Classical French, i.e., the seventeenth century. However, after a lifespan of some 600 years, it disappears rather quickly, though not entirely unexpectedly, during the first half of that century.
Ains is a more recent spelling mostly used after the Old French (OF) period; as usual, there are a few spelling variants of this word, the most common in addition to the s-form being ainz, mostly used in the older texts. There is also a longer variant ainçois, which has roughly the same meaning and the same functions as ains. However, the temporal function is the prevailing one with ainçois, which is also less common than ains (but cf. Buridant 2000: 561 and Marchello-Nizia 1997: 324 on the increasing popularity of ainçois in Middle French). As a conjunction, ains roughly corresponds to English but, but it also has temporal meanings as an adverb and as a preposition, meaning ‘before’ or ‘earlier’ (cf. section 4). The temporal uses relate to its etymology. Ains and ainçois derive from a hypothetical Latin *antius, which is the comparative of the temporal adverb and preposition ante, meaning ‘before’. Some of the uses of ains reflect this etymology, and the Italian conjunction/adverb anzi also has a related etymology (< antea) (see e.g., TLFi s.v. ains and Grevisse & Goosse 1993: 1553, 2008: 1399; cf. also note 9).
Ainz/ains appears in some of the earlier texts, but not in the first ones from the ninth century. One of the oldest examples ─ apparently the oldest one judging from the information retrievable from the current sources ─ is the following one from a hagiographic poem: 
Tuz l’escarnissent, si’l tenent pur bricun; (…)
Ne s’en corucet giens cil saintismes hom,
Ainz priet Deu quet il le lur parduinst
(La Vie de Saint Alexis, 11th century; Tobler & Lommatzsch vol. I: 54d) 
‘They all mock him and consider him a good-for-nothing
The holy man does not take offence
But prays God forgives them.’
Other examples from the Old French period, taken from Tobler & Lommatzsch (vol. I: 248):
Mes n’i areste ne demore,
Ainz s’an part an mout petit d’ore.
(Chrétien de Troyes, Chevalier au lion; ca. 1180)
‘But he does not stop nor stay,
But leaves very quickly.’
ne vaut plus targier,
Ains suit le cerf
(Guillaume le Clerc, Fergus; beginning of the 13th century)
‘He does not want to delay,
But follows the deer.’
There are few extant Old French texts until about the twelfth century, and since the Old French period extends from the ninth century to the beginning or middle of the fourteenth century, it covers about 500 years. Only about the 200 last years were a period of very intense literary activity (mid-twelfth to mid-fourteenth centuries). Middle French (MF) covers about 150 years, from the fourteenth century to the end of the fifteenth. The term ‘Renaissance French’, or simply sixteenth-century French, has recently been used to refer to the fairly short period between Middle French and Classical French, which extends roughly to the French revolution. (For the division of the French language into periods, see for example Picoche & Marchello-Nizia 2008: 186ff., Huchon 2002 or Rickard 1974.) After a period of blossoming in Old French, ains becomes much less common in Middle French. This is an often repeated claim (Rodríguez Somolinos 2002: 508, 516, van Reenen & Schøsler 2000: 96) which may not be based on solid statistics, but it is true of course that ains eventually disappears altogether. There is a clearer decline in the sixteenth century according to Rodríguez Somolinos (2002: 538; cf. also Buridant 2000: 561), and ains is practically not used any longer in the second half of the seventeenth century (Haase 1935: 390, Marchello-Nizia 1997: 325, 358; Picoche & Marchello-Nizia 2008: 311).
For this study I have collected some 150 examples from the sixteenth and early seventeenth centuries, which would of course need to be systematically compared with Middle French and Old French. The corpus comes mainly from the site of Project Gutenberg, where I have gone through all the texts from the sixteenth century and the first two thirds of the seventeenth. 
In addition to the question of the rather fast decline of the conjunction at the end of its lifespan, and its almost total disappearance within a few decades in the seventeenth century, another interesting question is the semantics and pragmatics of ains, which I will briefly present below. The first study dealing with the semantics and pragmatics of the conjunction is that of Melander (1916), followed by Antoine (1958) and Andersson (1965). Since Kleiber (1978), there have been several articles dealing with these questions (especially Rodríguez Somolinos 2000, 2002, van Reenen & Schøsler 2000, and Badiou-Monferran 2007). They have also been discussed in some histories of the French language, like Picoche & Marchello-Nizia (2008) (see also, for Middle French, Marchello-Nizia 1997). A useful summary of the main points and problems can be found in Buridant (2000), based on Kleiber (1978). All these studies basically agree on the premises presented below in section 3 – the distinction between ains and mais, their similarities and differences, and the existence of two different mais in French since the Middle Ages.
3. The semantics and pragmatics of ains
Ains is generally speaking an adversative conjunction, like mais, the equivalent of but. Mais appears in early texts (note 3), with roughly the same meaning or functions it has now; its principal function (the one corresponding to but, aber, ma, pero, men, etc.; see below) is still the same as it was more than 1000 years ago, and it has retained the same graphic form (see also note 14). A prototypical use of mais would be the one in examples (2a) and (2b):
Il est riche mais généreux.
He is rich but generous.’
Il n’est pas riche, mais il est généreux.
‘He is not rich but he is generous.’
In French linguistics this use of mais has been called “modifying” or “argumentative” (Badiou-Monferran 2007: 7; Buridant 2000: 558; Kleiber 1978, Rodríguez Somolinos 2000, 2002). The clause or phrase following the conjunction presents something which is not necessarily incompatible with the first clause, but which gives somehow unexpected or surprising, perhaps apparently contradictory additional information. There is also another use of mais, “mais2” (Badiou-Monferran 2007: 8; Buridant 2000: 559; Kleiber 1978, Rodríguez Somolinos 2000, 2002):
Il n’est pas riche, mais pauvre.
‘He is not rich, but poor.’
This use of mais has been called “refutative”, “corrective” or “exclusive”. Here mais indicates incompatibility; the phrase or clause following the conjunction refutes the positive assertion contained in the first, negative clause, or corrects it. This brief description will suffice here, since, as will be mentioned in the next paragraph, this distinction exists in several languages belonging to at least two language families, and, on the other hand, the functions of the connectors have been formalized for example within French pragmatics. 
The distinction between the two types of mais in examples (2) and (3) can be found in other languages, namely Finnish (mutta/vaan), Swedish (men/utan), German (aber/sondern), Italian (ma/anzi) and Spanish (pero/sino).  But strangely this lexical distinction exists neither in Modern French nor in English.  There are interesting similarities between the behaviour of these pairs in the languages mentioned (even though it is not surprising to discover differences too), and they would deserve a study of their own.  As for English, but may not always be the best choice in English to translate ains or Modern French mais. In some cases one could probably use on the contrary, or leave out the connector altogether. (See also the suggested translations of examples (4c), (7a), (8) and (9) below; they offer the possibility of using in English a translation other than but).
I will briefly describe below some of the salient characteristics of the conjunction, which it also shares in part with the equivalents in the modern languages mentioned above. These include negation in the first clause (3.1), the possibility of ellipsis (fairly frequently) (3.2), and the use of the same subject in the second clause (most frequently, though the number of exceptions is growing over time according to my corpus) (3.3). In addition, V-S inversion (3.4) was possible in Medieval French, but disappeared later and is not present in the modern languages mentioned above. (However, the structure of several sentences does not allow any possibility of inversion, e.g., when there is no subject in the second clause.) I will end section 3 with some considerations on the possible renderings of ains in other languages (3.5).
3.1 Negation in the first clause
What is common to the conditions of use of the second term (Modern French “mais2”) of each of the above pairs in all languages (Finnish (mutta/vaan), Swedish (men/utan), German (aber/sondern), Italian (ma/anzi) and Spanish (pero/sino)) is that it must be preceded by negation, generally explicit but also implicit in some cases. Examples without overt negation are rather few in French. In some cases, it may be enough to have for example a negative prefix in the first clause (example (4c) from the sixteenth century):
Tout ce que j’ay dict icy, ne prejudicie en rien aux saiges ains les rend de tant plus illustres
(Recueil des exemples de la malice des femmes, et des malheurs venus à leur occasion; 1596)
‘What I have said here will not harm at all the wise men, but will make them even more illustrious.’
Bestes mues ne le font pas. / Ains elles ament et cherissent / Et suivent toutes pas a pas / Leurs meres de qui elles yssent
(Le chevalier des dames du dolent fortuné; 16th century)
‘Mute animals do not do it / but they love and cherish / and follow every step / of their mother from whom they come.’
et que changez conditions et ayez meilleures manieres que par avant qui sont des plaisantes au monde / ains serez doulx courtois: humain:
(La terrible et merveilleuse vie de Robert le Diable; 1530)
‘and you must change your way of life and improve your manners which are unpleasant to the world / rather you must be gentle, polite, human.’ 
There are a few examples with no negation whatsoever. This may be due to idiolectal or dialectal variation (cf. some uses of vaan in Finnish). 
De cela soyés tout conclus
Ains mon entente est de mourir
Ceans cordelier et reclus
(Martial d’Auvergne, L'amant rendu cordelier a l'observance d'amour; 2nd half of the 16th century)
‘You can be assured of that
Rather my intention is to die
Here as a monk and hermit.’
3.2 Finite verb vs. ellipsis in the second clause
Whether the second clause must contain a finite verb or not, or whether both are possible, seems to be language-specific (Italian requires a finite verb, Spanish does not allow one; cf. note 8). In Old French, a finite clause after ains was required. In example (1) above, this is the case, since the focus of negation is the verb (he does not take offence, but prays to God). Elliptical clauses became common only in Middle French. Examples (6a) and (6b) are more recent (dating from the sixteenth and seventeenth centuries):
Pas n'y viens par joyeux confort
Ains pour mener vie solitaire
(Martial d’Auvergne; 2nd half of the 16th century)
‘I am not going there for pleasure
But in order to lead a solitary life.’
Il y avoit plusieurs de nos sauvages qui ne vouloient point ceste guerre, ains la continuation de la paix avec les Yrocois
(Samuel Champlain [travel account]; after 1630)
‘There were several savages of ours who did not want this war, but the continuation of peace with the Iroquois.’
Example (6b) is a rather late one, from the fourth decade of the seventeenth century. As pointed out earlier, ains practically disappears around the middle of the seventeenth century. The explorer Champlain, the founder of the city of Quebec in Canada, uses the conjunction frequently in his travel accounts during the first half of the century. One might venture to suggest that this is due to the fact that because he was constantly abroad, he lost touch with the contemporary French language.
3.3 Same subject in both clauses
Another feature of ains/mais2 is that both clauses most often have the same subject. There are exceptions, however, which become more frequent with the passing of time. Example (7a) illustrates the rule and example (7b) the exception (both from the sixteenth century):
Mais tu jureras
Que point ne _ [ellipted subject] diras
Que rien dit je t’aye
Ains _ [ellipted subject] affermeras
Que rien n’en sçavoie
(La terrible et merveilleuse vie de Robert le Diable; 1530)
‘But you will swear
that you will not say
that I have told you anything;
on the contrary you will assure
and (you will) declare
that I did not know anything about it.’
A quoy _ [ellipted subject] respondirent faisant leur excuse quilz ne luy avoient pas conseille sen aller ains quelle sen estoit allee parce que les paiges lavoient battue
(Jacques Cartier, Voyage au Canada; 1st half of the 16th century)
‘To which they replied that they had not told her to go away, but she had gone because the pages had beaten her.’
Verb-subject inversion was possible in Medieval French, but is not found in any of the languages mentioned above. However, since the subject (of the second clause or both clauses) is still frequently left out in the sixteenth century, inversion seldom materializes. This seems to be by far the most frequent case; in (8a), the pronoun of the 1st person singular is missing altogether and the verbal form croy is still sufficient to refer to the right person (but other forms of the first person do appear in this example: m’, mes); in (8b) the subject of the first clause, mon filz, is not repeated in the second one. (8c) presents a case of inversion, while (8d) retains the henceforth canonical S-V order.
Le conseil de dame venus
Ne m’a point la douleur ostee
De qui mes membres sont tenus
Ains _ [ellipted subject] croy qu’elle soit augmentee
(Anonymous, Livre d’amours; 16th century)
The advice of Venus
Did not remove the pain
Which tortures my limbs
On the contrary I think it has increased.’
Car tant que vous serez en ce monde, mon filz [subject] au moins n’aura point faulte de mere: ains _ [ellipted subject] sera eslevé et nourry par vous
(Mellin de Saint-Gelais, Sophonisba; middle of the 16th century)
‘Since as long as you are alive, my son will not lack a mother; rather he will be raised and fostered by you.’
Et nous arrivans a l’ung des boutz dudict lac, ne nous apparoissoit aucun passaige n’y sortie: Ains sembloit [verb] icelluy [subject] estre tout cloz sans aucune riviere,
(Jacques Cartier, Voyage au Canada; 1st half of the 16th century)
‘And when we arrived at one end of the lake, we did not see any passage nor exit: on the contrary it [the lake] seemed to be entirely closed without any watercourse.’
Cela fust resolu, mais leurs ennemis ne firent pas ce que l’on avoit pensé: ains ils [subject] se préparèrent toute la nuict,
(Samuel Champlain [travel account]; ca. 1600)
‘This was decided, but their enemies did not do what was expected; rather they made preparations all night.’
3.5 Equivalents of ains in other languages
If one reflects on how to render ains in the above examples in Modern French, one easily comes to the conclusion that mais is not the best possible choice. Possible translations could be au contraire or plutôt. But contemporary French does not really seem to need any separate or independent equivalent to ains, no more than English. It is not obvious at all that mais is the best or most natural choice in all cases where one would use sondern in German or anzi in Italian. For example, to render in French the Italian sentence: Non è ricco, anzi è povero, one could imagine that in spoken language there would be no conjunction at all: Il n’est pas riche, il est pauvre! (‘He is not rich, he’s poor!), or mais, as a kind of pragmatic marker, could be in sentence-initial position: Mais il n’est pas riche, il est pauvre! (‘But he isn’t rich, he’s poor!’)  However, dealing with Modern French examples is well beyond the scope of this paper, and I do not have a relevant corpus for that.
On the other hand, it is not obvious that in Medieval French ains should always be considered as indicating refutation or correction. Consider this example from the thirteenth century (OF):
Ostrisses est une grant beste ki a eles et plumes a samblance d’oisel, et piés comme chamel ; et ne vole pas, ains est griés et pesans par sa complexion, ki est si oublieuse malement k’il ne li sovient des choses passees.
(Brunet[to] Latin[i] ; A. Henry, Chrestomathie de la littérature en ancien français, 1973: 329)
‘An ostrich is a big animal which has wings and feathers like a bird, and feet like a camel; and [it] does not fly, rather/since it is heavy and its body weighs a lot, which [=and it] is so terribly forgetful that it cannot remember past things.’
Here it would be logical to think that the second clause gives an explanation to the first one, rather than indicating a correction or a refutation. Though there is apparently nothing un-French in this sentence, one can wonder whether Brunetto Latini may not have been influenced by his native Florentine or Tuscan and the form anzi, which behaves somewhat differently from ains, but also functions as a conjunction, an adverb and a preposition. Since ains and anzi are etymologically related (contrary to Spanish sino), a comparison of their development might be in order; this could very well also include Spanish.
4. The decline and disappearance of ains
The rather fast disappearance of ains is intriguing (pace Badiou-Monferran 2007). Its death was regretted by some authors after 1650 (e.g., Dupleix 1651 and La Bruyère (second half of the seventeenth century)). It had still been used during the first half of the century by certain authors, like François de Sales (1567–1622) and Scarron (1610–1660).  The most recent example found in my corpus is from the 1630s (example 6b).  However, the use of ains was criticized as obsolete during the same period, before 1650, by eminent grammarians and authorities on language like Malherbe (1606) and Vaugelas (1647) (note 12).
This disappearance has been attributed to the confusion of uses between ains and mais (e.g., Rodríguez Somolinos 2000, 2002), and to the latter taking over the refutative function of ains. Somehow this leaves open the question of why French did not retain this handy lexical pair found in several languages. Badiou-Monferran bases her article at the outset on Kleiber (1978); this is the first linguistic study on ains, which is called an “opérateur d’inversion” by Kleiber (Badiou-Monferran 2007: 9). Badiou-Monferran goes through the various types of explanations given and classifies them into three groups. She rejects the ”semantic-logical” explanation proposed by Buridant (2000) and the “micro-syntactic” one represented by Melander (1916), the first one to have dedicated a study to the problem, followed by Antoine (1958). I have some difficulties in following Badiou-Monferran’s reasoning throughout her article, but she objects to Buridant’s claim (2000: 561) according to which the disappearance of ains is related to the sixteenth-century craving for precision and the attempt to avoid ‘tautology’. She may indeed be right in disagreeing with Buridant’s ‘conventional’ explanation, but her arguments seem less convincing. She claims that there is an extensive rivalry between ains and mais2 in the sixteenth century (Badiou-Monferran ibid.: 11). In fact this rivalry already exists in Old French (see examples of mais in Tobler & Lommatzsch, vol. I: 863–864). She asserts that ‘ains is no more perceived by French speakers and writers as an inversion operator’ (ibid.: 12–13), rather a strong claim, and also hard to prove.
Her own “macro-functional explanation” involves indeed a complicated, multi-phase process. It is based on the grammaticalization studies of Marchello-Nizia (cf. now especially 2009), and is very intricate, which makes it hard to prove or to refute. She bases it on statistical evidence (see note 13) which is interesting but scarce, and refers to the instability of ains and its categorial fuzziness, in connection with the adverbial phrase ains que ‘rather than’ (ibid.: 16–18). She seems to associate categorial fuzziness both with ains and ainçois, claiming that the latter is the cause of the disappearance of the former, which in turn triggers the grammaticalization of au contraire (Badiou-Monferran 2007: 19–20; see also below). She concludes that her analysis helps explain the fact that ains took over, just before disappearing, the functions of mais2, the order in which the various functions of ains disappeared, and the above-mentioned grammaticalization of au contraire.
I would rather suggest that the multiple functions of ains in Old French and still in Middle French might help us find an explanation. Multifunctionality was a typical feature of Old French, which eventually resulted either in the decrease of functions  or in the disappearance of the lexeme altogether (cf. note 17). Ains was indeed a temporal adverb (‘before, earlier’) and preposition (‘before’), but also, as we have seen, an adversative conjunction in Medieval French. It could also be used in the conjunctive phrase ains que ‘before’. In addition, ains appeared in some idiomatic expressions like ainz ne:
Ainz ne verrat passer cest premier meis
Que je l siurai od mil de mes fedeilz
(Chanson de Roland, ca. 1100; Tobler & Lommatzsch vol. I: 247)
‘He will not see this first month end
before I will follow him with 1000 men of mine’ [Literally roughly: ‘before he will not see this first month end than/when I will follow him …’]
or as a kind of sentence-initial  pragmatic or discourse marker as in
Ja ainz n’iert vespres ne li solauz couchans,
Ja la verrai ardoir en feu ardant
(Ami et Amile, 13th century; Picoche & Marchello-Nizia 2008: 351)
‘Before tonight and sunset
I will see her burning at the stake’
[Literally roughly: ‘earlier it will not be night nor sunset than/when I will see her burn …’]
There was also the prepositional phrase ainz de ‘rather than, instead of’, indicating preference, but also opposition, in addition to the temporal meaning ‘before’. See Huguet (vol. I: 140–141), Picoche & Marchello-Nizia (2008: 301, 312, 325, 331, 351) and Tobler & Lommatzsch (vol. I: 247–249).
The semantic-pragmatic development of ains can be described with the sequence anteriority → preference → opposition (Kleiber 1978, Buridant 2000; see note 9). It seems that it is also roughly in this order that its uses in Medieval French started to fade out. Instead of the grammaticalization referred to by Badiou-Monferran, one might allege here a pragmatic development (cf. Marchello-Nizia 2009, Traugott & Dasher 2002 and van Reenen & Schøsler 2000). According to Badiou-Monferran, ains is often accompanied by the “adverbial inverter” au contraire ‘on the contrary’ (Badiou-Monferran 2007: 12), but her examples are rather scarce and it remains to be proved whether this was possible earlier too. She ties the disappeance of ains to the grammaticalization of plutôt (← plus tôt; ‘earlier’ → ‘rather’, a process which ended as late as the sixteenth century) and au contraire. One can wonder whether the latter phrase can indeed be considered as grammaticalized, but in any case ains seems to have incurred a pragmatic development, which ended in the adversative or refutative use prevailing in early Modern French. It might be suggested that since the various other uses of ains had become outdated before the pragmatic function did, the form was felt as antiquated altogether. But then one might wonder why the temporal uses of ains disappeared; the old traditional preposition and adverb avant, less ambivalent or ambiguous than ains, may have been felt to have some advantages over the latter. In any case, it does not seem obvious to me that clear-cut distinctions should or need be made between for example allegedly syntactic, functional or pragmatic explanations.
5. Concluding remarks
The multiple, in fact too various and numerous, functions of ains, along with its rivalry with mais, can contribute to explaining why the in principle handy adversative function did not suffice to allow ains to survive, contrary to the other Romance and Germanic languages referred to above. In addition to the problematic multifunctionality, phonetic attrition is not to be excluded. The pronunciation of the connector must have been reduced to just one nasal phoneme after the Middle French period at the latest, even though the Trésor de la langue française (TLFi) mentions the pronunciation of the final s.  French is of course a language known for its numerous monosyllables, including the conjunctions et and ou, the pronunciation of which consists of just one phoneme. Due to their frequency and contexts of use, they are not problematic in spite of their shortness, but one can imagine that ains might be, if it were used in present-day French. 
 When possible, I will use here the traditional term “conjunction”, which appears in the relevant literature in French along with “connector” (“connecteur”). In French there is no equivalent to the English “connective”. I assume that there is no unanimity on the distinction “connective/connector” in English. According to the French practice, I will use “connector” when it seems more adequate than “conjunction”, e.g., when referring to a phrase such as “on the contrary”. Buridant (2000) uses the term “conjonction” for mais and ains, while Rodríguez Somolinos (2000, 2002) and Badiou-Monferran (2007) speak of “connecteurs”. Van Reenen & Schøsler (2000) use “particle”, which is of course more general.
 The English translations of the examples are mine. – I am grateful to the reviewers of a previous version of this paper for their comments.
 The first occurrence of the conjunction mais in the same adversative or refutative function (see section 3) also appears in this text (see Tobler & Lommatzsch vol. V: 863).
 120 examples come from the Gutenberg site (the examples given here are from this source, if not otherwise stated), 23 from two anthologies of sixteenth- and seventeenth-century texts edited by Peter Rickard, and a few from my own corpus of late sixteenth- and early seventeenth-century texts collected for other purposes. I have naturally excluded examples where ains has another, for example, a clearly temporal meaning. – The Project Gutenberg site allows free downloading of digitized books in several languages and different e-forms; the transcriptions are diplomatic and in several cases would require an annotated edition to make the text clearer. For research purposes, the site is particularly interesting, seeing it makes available printed texts from the period subsequent to the invention of printing (mainly the sixteenth century), often difficult to find nowadays in printed form. Sixteenth-century examples can also be found in Huguet (1925–1967).
 Some of these pairs are mentioned in previous studies, such as Buridant (2000: 561) or Kleiber (1978: 271), but the list of languages presented is the most complete one I know of (at least Finnish and Swedish are not mentioned in the studies I am familiar with). I am grateful to my colleagues Eva Havu, Elina Suomela and Begoña Sanromán for information on German, Italian and Spanish respectively.
 English has never had a connector corresponding to the adversative or rectifying use of ains discussed here (information kindly provided to me by Leena Kahlas-Tarkka).
 E.g., in Italian, contrary to the other languages mentioned above, the clause starting with anzi cannot be elliptical: Non è ricco, anzi è povero ‘He is not rich, on the contrary he is poor’. Notice that Finnish aligns itself with the other languages mentioned, though it is not an Indo-European language but a Finno-Ugric one.
 Compare the development of ains with that of rather, which also evolves from a comparative and a temporal adverb (‘more quickly, sooner’) to one expressing priority, preference, contrast, etc. (OED online, s.v. rather adv. 5a). Ursula Lenker points out to me the parallel with German eher. This development also concerns Italian anzi (see section 1). Cf. also Rissanen (1999).
 Cf. this example from a magazine (Suomen Kuvalehti 2010): Ylimyksen pystykuva pysyy ja pölyttyy kansakunnan kaapin ylähyllyllä, vaan samalla monet taiteilijat ja kirjailijat ovat surutta kolhineet suurmiesmyyttiä. ‘The bust of the aristocrat is gathering dust on the top of the cabinet of the [Finnish] nation, but at the same time several artists and writers have dented without restraint the myth of the great man.’ From the point of view of the standard language, the choice of vaan instead of mutta is contrary to usage and to the rules of grammar, mainly because the first clause is affirmative and not negated. But cf. also Hakulinen & al. (2004: 1052) about the interchangeability of mutta and vaan in certain contexts.
 These two coined French examples would probably reflect oral usage in the non-omission of the verb of the second clause, whereas (3) sounds more formal. There would also be some relevant prosodic features, like emphasis on the two adjectives riche and pauvre.
 Badiou-Monferran (2007) has investigated the Frantext database and found a few last examples from the period 1650–1659. Frantext is a database allowing the search of hundreds of texts from the Middle Ages to the present day, but it can only be used by subscription.
 Mais is a case in point, since most of its medieval functions have disappeared in the course of time, and what remains is its use as a co-ordinating conjunction. As an adverb, it was used with the meanings ‘(any) more’, ‘ever’, ‘now, henceforth’, etc., as a conjunction (‘because’) and in several combinations as an adverb (mais que: ‘except’, ‘though’, ‘provided that’, ‘namely’, etc.). See the impressive amount of examples given in Tobler & Lommatzsch vol. V: 855–875.
 The temporal adverb ja (cf. Modern French jamais) starting the sentence does not really invalidate the sentence-initial status of ainz; ja has here a mainly pragmatic function and could be omitted, if it were not for the needs of metre (decasyllable). On the other hand, ja and ainz also do appear together as a pragmatic cluster.
 The fact that ains is mentioned at all in this dictionary, which concentrates on Modern French since the Revolution, is due to its totally anachronistic use by a few 19th-century authors. If one were to use ains in speech, it would be logical to pronounce the final s, because the particle would otherwise sound too reduced and could create comprehension problems (it would be homonymous with the current colloquial interjection and pragmatic particle hein and with the numeral and indefinite article un).
 In any case, quite a few Old French particles disappeared well before the seventeenth century, and one would be hard-put to find valid explanations for all of them. See for example Picoche & Marchello-Nizia (2008: 301–324): atot, aval, coste, empur, endroit, enz, joste, lez, emprés, adés, eneslepas, lués, onques, pieça, onques, céans, iluec, leans, etc. (this is just a small selection).
Project Gutenberg. http://www.gutenberg.org/.
Henry, Albert. 1973. Chrestomathie de la littérature en ancien français. Bern: A. Francke.
Huguet = Huguet, Edmond. 1925–1967. Dictionnaire de la langue française du seizième siècle. 7 vols. Paris: Champion & Didier.
OED = Oxford English Dictionary. 2010. http://www.oed.com/.
Rickard, Peter. 1992. The French language in the 17th century. Woodbridge: Brewer.
Rickard, Peter. 1968. La langue française au seizième siècle: étude suivie de textes. Cambridge: Cambridge UP.
Tobler & Lommatzsch = Tobler, Adolf & Lommatzsch, Erhard. 1915–. Altfranzösisches Wörterbuch. Berlin, Wiesbaden, Stuttgart: Steiner.
TLFi = Le Trésor de la langue française informatisé. Nancy: CNRS–ATILF. http://atilf.atilf.fr/.
Andersson, Sven. 1965. “L’opposition mais/ainz (ainçois)”. Studia Neophilologica 37: 40–44.
Antoine, Gérald. 1958. La coordination en français. 2 vols. Paris: Editions d’Artrey.
Badiou-Monferran, Claire. 2007. “Disparition de ains et évolution du système grammatical”. Études sur le changement linguistique en français, ed. by Bernard Combettes & Christiane Marchello-Nizia, 7–26. Nancy: Presses universitaires de Nancy.
Buridant, Claude. 2000. Grammaire nouvelle de l’ancien français. Paris: Sedes.
Charaudeau, Patrick & Dominique Maingueneau. 2002. Dictionnaire d’analyse du discours. Paris: Seuil.
Ducrot, Oswald & Carlos Vogt. 1979. “De magis à mais: une hypothèse sémantique”. Revue de linguistique romane 43: 317–341.
Grevisse, Maurice & André Goosse. 1993/2008. Le bon usage. 13th/14th edition. Bruxelles: De Boeck – Duculot.
Haase, Albert. 1935. Syntaxe française du XVIIe siècle. (Nouvelle édition traduite et remaniée par M. Obert.) Paris: Delagrave.
Hakulinen, Auli, Maria Vilkuna, Riitta Korhonen, Vesa Koivosto, Tarja Riitta Heinonen & Irja Alho, eds. 2004. Iso suomen kielioppi [Big Finnish grammar.] Helsinki: Suomalaisen kirjallisuuden seura. http://kaino.kotus.fi/visk/etusivu.php.
Huchon, Mireille. 2002. Histoire de la langue française. Paris: Le livre de poche.
Kleiber, Georges. 1978. “Sur l’emploi adversatif de mais et de ainz (ainçois) en ancien français”. Travaux de linguistique et de littérature 16: 271–292.
Marchello-Nizia, Christiane. 1997. La langue française aux XIVe et XVe siècles. Paris: Nathan.
Marchello-Nizia, Christiane. 2009. Grammaticalisation et changement linguistique. (=Collection Champs linguistiques.) Bruxelles: De Boeck – Duculot. [E-book available]
Melander, J. 1916. Étude sur magis et les expressions adversatives dans les langues romanes. Uppsala: Almqvist & Wiksell.
Picoche, Jacqueline & Christiane Marchello-Nizia. 2008. Histoire de la langue française. Paris: Éditions Vigdor. [E-book, available for free.] http://www.lexilogos.com/francais_langue_histoire.htm.
Reenen, Pieter van & Lene Schøsler. 2000. “The pragmatic functions of the Old French particles ainz, apres, donc, lors, or, puis, and si”. Textual Parameters in Older Languages (=Current Issues in Linguistic Theory, 195), ed. by Susan Herring, Pieter van Reenen & Lene Schøsler, 59–105. Amsterdam/Philadelphia: Benjamins.
Rickard, Peter. 1974. A History of the French Language. London, etc.: Hutchinson.
Rissanen, Matti. 1999. “On the adverbialization of RATHER: Surfing for historical data”. Out of Corpora: Studies in Honour of Stig Johansson. (=Language and Computers, 26), ed. by Hilde Hasselgård & Signe Oksefjell, 49–59. Amsterdam/Atlanta: Rodopi.
Rodríguez Somolinos, Amalia. 2000. “Mais, ains, ainçois en moyen français: syntaxe et sémantique”. Le moyen français 46–47: 449–465.
Rodríguez Somolinos, Amalia. 2002. “Ainz et mais en ancien français”. Romania 120: 505-541.
Rossari, Corinne, Anne Beaulieu-Masson, Corina Cojocariu & Anna Razgouliaeva. 2004. Autour des connecteurs. Réflexions sur l’énonciation et la portée. Bern: Peter Lang.
Spillebout, Gabriel. 1985. Grammaire de la langue française du XVIIe siècle. Paris: Picard.
Traugott, Elizabeth Closs & Richard B. Dasher. 2002. Regularity in Semantic Change. Cambridge: Cambridge University Press.