Les Chroniques de langage and the Development of Linguistic Purism in Québec

The second half of the twentieth century has seen the emergence in Quebec of a number of societies for the protection or promotion of the French language. Since the 1990s, new societies have increasingly been established online, with older ones also creating an online presence. Most of these societies publish journals which appear several times a year. An earlier study (Walsh 2013) has shown that the type of metalanguage used and the major preoccupations reflected by these societies on their websites and in their journals can be seen to reflect a moderately purist attitude (based on the theoretical framework for evaluating and measuring purism outlined by George Thomas 1991). This article presents the results of a similar investigation of a sample of chroniques de langage, columns dealing with questions of language, which appear regularly in journals and newspapers. The content and metalanguage of chroniques from two periods (1880–1889 and 1940–1949) will be compared, to determine whether the current-day ...


Introduction
The second half of the twentieth century saw the emergence in Quebec of a number of societies for the protection or promotion of the French language. Since the 1990s, new societies have been increasingly established online, with older ones also creating an online presence. Most of these societies publish journals which appear several times a year. An earlier study (Walsh 2013) 3 has shown that the type of metalanguage used and the major preoccupations reflected by these societies on their websites and in their journals reflect a purist attitude. This article presents the results of a similar investigation of chroniques de langue -newspaper columns dealing with questions of language which appear regularly in journals and newspapers -from the same region but an earlier time period. We may expect this sample of language columns from Quebec to be similar in nature to the material produced by the language societies in Quebec: both are the result of a perception that the language is in some way endangered and a desire to protect the language from that danger.
We may also expect differences in the language used and the concerns displayed in the earlier period, given the immense changes in society and in attitudes towards the French language which were brought about by societal change and linguistic legislation from the 1960s on.
The study analyses a sample of 200 articles from the ChroQué corpus of language columns, which contains over 8000 articles produced by forty columnists over the period 4 taking one hundred texts from two decades at different periods (1880-89 and 1940-49). In particular, the type of metalanguage used and the major concerns of the authors are examined during the course of this article to determine similarities with, and differences from, the language and concerns of current-day language societies. Before proceeding to a discussion of the study, it is first necessary to outline what we understand by the term 'linguistic purism' and the methods used to measure it. The analysis in this article uses the theoretical framework developed in George Thomas's Linguistic Purism (1991) 5 , 'the first broadly comparative and cross-cultural study of purism', which establishes a theoretical framework outlining the elements typical of all purist movements. 6 This framework is in the form of a checklist of characteristics, presented in a questionnaire, by means of which he claims we can complete a profile of purism in a given country. 7 The framework was used in Walsh's (2013) 8 study of the objectives, attitudes, major preoccupations and type of language used by six language societies for the defence or promotion of the French language in Quebec and showed that these societies display a moderately purist attitude. It is applied here to the sample of language columns to determine whether they also reflect a purist outlook.
This article takes the broad definition of purism outlined by Thomas: 9 the manifestation of a desire on the part of the speech community (or some section of it) to preserve a language from, or rid it of, putative foreign elements held to be undesirable (including those originating in dialects, sociolects and styles of the same language). It may be directed at all linguistic levels, but primarily the lexicon. Above all, purism is an aspect of the codification, cultivation and planning of standard languages.
According to this definition, purism includes ideas which are also common to standardisation and prescriptivism, namely the notion that some elements of language are desirable while others are not. However, purism goes further by introducing the following ideas: only one form of the language is the 'correct' form (generally this form is the standard form); this language is currently pure and, therefore, change to it equals contamination, corruption or decline of some sort; and the language must be protected from this contamination and preserved in its current state (or, alternatively, if the language has already begun to be corrupted, the corrupted part must be removed). Standardisation and prescriptivism do not necessarily share these themes of contamination, protection and preservation.
Standardisation, for example, frequently requires change and innovation, particularly during the elaboration stage, and prescriptivism, although it shares with purism the idea that there is only one 'correct' form of the language, is not inherently conservative; prescriptive norms can change.
The article first outlines the context for the study and the sources used. It next provides an analysis of the main preoccupations of, and the metalanguage used in, the sample of language columns chosen and compares this to the preoccupations and metalanguage of the language studies analysed in Walsh 2013. 10 It then applies Thomas's framework to the results of this analysis to determine whether or not the language columns can be seen to display a purist attitude. Finally, it proceeds to a discussion of the outcomes of this analysis and the earlier one to determine where similarities and differences lie and to discuss possible reasons for any changes over time.

Context for the study and sources used
Walsh (2013) 11 used language societies as a way of accessing attitudes towards the French language in the present day to determine whether or not they displayed linguistic purism. The aim of the current study is to assess attitudes towards French in the past using the strong tradition of chroniques de langage, language columns in the periodical press dealing with questions of language, which have been published in Quebec since the 1860s. More specifically, chroniques de langage are articles discussing questions relating to language which are produced by a single author and published regularly, generally in the written press (in newspapers and journals). The content of such columns can be very varied, both across different columns/authors and also within a single column. According to Remysen, 12 who maintains the Québécois corpus of chroniques de langage, ChroQué (from which the sample of articles for this study is taken) and who has published extensively on the topic: La chronique de langue est un ensemble de discours sur la langue, plus particulièrement encore sur les bons et les mauvais usages de la langue. Elle est diffusée périodiquement sous forme de rubriques dans les médias écrits (articles de journal ou de revue) ou électroniques (émissions de radio ou de télévision). La chronique est signée par une même personne, physique ou morale, à laquelle on reconnaît une compétence en matière de langue.' These language columns were first studied in the 1970s and they have been used since then to examine various aspects of linguistic attitudes in Quebec such as attitudes towards anglicisms (cf. Bouchard 1989), 13 linguistic insecurity (cf. Bouchard 1988) 14 and, more recently, the 'imaginaire linguistique' and the links between language and identity (cf. Wim Remysen 2011; 15 2012). 16 Studies of the output of individual authors of these articles have also been published (for example, Gagné 2005; 17 Laurendeau 2004; 18 Prévost 1996). 19 These articles have not yet been used specifically for a study of linguistic purism, but I suggest that they are particularly useful documents for a study of the development of purism in the past, as the tradition of chroniques dates back to at least the mid-19th century, a high number of columns were produced which reached a wide audience 20 and we know that the attitudes of the intellectual elite who produce them have an important influence on general attitudes towards language. 21 The first metalinguistic discussion of Canadian French 22 can be seen in the chroniques de langue published by Michel Bibaud in the journal L'Aurore from 1871, but it was not until 1841, with the publication by Thomas Maguire of the Manuel des difficultés les plus communes de la langue française 23 that a metalinguistic discourse really began to emerge. 24 Bibaud's language columns consisted of short articles condemning the use of anglicisms in specialized vocabulary, in particular legal vocabulary, but the language used by the French speech community more generally in Quebec was not attacked. However, Maguire's Manuel was a corrective volume, aiming to improve the French used generally in Canada, and was based solely upon the French standard as used in France, not accepting any term that was not included in French dictionaries. From this point on, a number of corrective manuals were published, most of which had the intent of removing 'barbarismes', 'expressions vicieuses' and 'fautes' from Canadian French, and a fashion for language columns also began. Between 1865 and 1960, 65 language columns appeared in French-Canadian newspapers, such as La Presse, La Patrie and Le Droit. These were nearly all concerned with denouncing 'to different degrees the terrible affliction of anglicisms' 25 and also targeted terms specific to Canadian French such as archaisms, regionalisms or neologisms. 26 For a detailed analysis of the work of various language columnists, see Wim Remysen 2010Remysen , 27 2011Remysen , 28 2012 In order to provide a representative sample of language columns over the period under examination, 100 individual articles have been taken for two ten-year periods between 1880-89 and 1940-49 from a sample of language columns written by four different authors. 1880-89 was chosen as the first decade, as this was the decade when the production of language columns first really became widespread. 30 1940-1949 was chosen as the second decade, because it was in the decades after this time that the 'Quiet Revolution' began to take hold (in particular the 1960s) and linguistic legislation began to be enacted (most importantly in the 1970s) that so changed attitudes towards the French language in Quebec (cf.  for a sociolinguistic history of Quebec). 31 Samples of articles from two language societies for the defence and promotion of the protecting the French language from anglicisms or 'incorrect' use. In this latter aim, they display a similar purpose to the language columns which are the focus of the present study.

1880-1889
A sample of articles from the language columns of two authors was taken for the period 1880-1889: Alphonse Lusignan and an anonymous author (see appendix for further details). 33 Lusignan's articles are briefer and less detailed than Anonymous; the former deals with only one issue per article, while the latter generally discusses a number of issues. 34 Both columns deal mainly with errors in the use of terms or expressions in French. Although both also frequently refer to the use of anglicisms, this remains a side issue to some extent. Of 402 individual comments on various issues, Anonymous mentions anglicisms 58 times, that is in only one in seven of his comments (14.4%). Although Lusignan mentions them more frequently, they appear in only 18 of 50 comments, that is, in roughly one-third of comments (36%). The remaining comments all deal with perceived errors in French, and tend to centre on the incorrect use of individual terms or phrases.

Anonymous
In Anonymous' column, the majority of comments focus on the perceived incorrect use of individual terms and phrases in French. For example, in several comments, Anonymous discusses what he sees as the incorrect use of 'au delà': 'Ne dites pas : au delà de 3000 hommes ; mais : plus de 3000 hommes.' (31 December 1880). Other comments discuss the correct use of the terms faire and infliger and the difference between parmi and dans. Anonymous frequently uses overtly prescriptive forms of language in his comments, including forms such as 'Ne dites pas X, mais Y'; 'Ne parlez pas de X, parler de Y'; and 'À l'expression X, préférez Y'. This makes it clear to readers that only one form is the correct form, and any other forms are to be avoided. In spite of this prescriptive stance, Anonymous rarely includes any other commentary carrying a value judgment. He occasionally reinforces the use of the prescriptive imperative 'ne dites pas' by stating that a form is incorrect or redundant or that it 'sounds bad', for example, 'La première forme est incorrecte' (7 January 1881) or 'La première expression ... sonne d'une manière peu gracieuse' (11 March 1881).
However, he only very rarely makes any clearly negative value judgements and does not make use of any imagery associated with purist discourse in other contexts (e.g. choking weeds, disease, contagion etc.).
Anonymous occasionally states that a certain term or expression 'n'est pas français', and as this can indicate both anglicisms and other French 'errors', itt would appear that it can mean both 'incorrect' French and literally 'not' French (i.e. foreign), although it most often means the latter. However, in general, when discussing anglicisms or usages influenced by English (in particular loan translations and semantic loans), Anonymous makes no reference to foreign influence and does not indicate that it is because of its status as an anglicism that the term or expression is to be avoided. Indeed, of the 58 comments which deal with anglicisms, only ten overtly refer to the foreign nature of the term or expression in question, namely by stating in each case that the usage 'n'est pas français'. He does not state what he means by 'français', and although one may assume that it means the standard French of France, he does not explicitly state this at any stage, nor does he mention any dictionaries or grammars.

LUSIGNAN Alphonse
Lusignan is less overtly prescriptive than Anonymous and uses fewer expressions of the type 'Dites ... ne dites pas', although expressions such as 'il fallait dire', 'il faut dire' and 'on a tort de dire' appear in just over two-fifths of the comments examined (44%). He otherwise presents his articles simply and without negative value judgments, only explicitly referring to something being a 'faute' on three occasions. Similarly to Anonymous, those comments dealing with perceived errors in Canadian French tend to focus on individual terms and phrases in French, for example, the correct use of the terms 'jalousies' and 'persiennes' (6 October 1884) or the difference between 'résidence' and 'demeure'/'domicile' (11 November 1884). He occasionally touches upon issues such as gender, pronunciation or spelling but does not go into any level of detail in any of his articles, and does not touch on more complex points of grammar at any point. He is more overtly negative towards anglicisms than Anonymous, however, , in particular, loan translations and semantic extensions. He employs more clearly negative language to describe their use, for example, 'un anglicisme féroce' (16 May 1884) or 'voilà un affreux anglicisme' (1 August 1884).
Lusignan is also unclear about the standard he is adhering to. He alludes to the French of France only once ('en France on emploie le mot exposition de préférence', 9 April 1885), and at no stage does he discuss the standard language or directly allude to the variety of French he is recommending to readers. However, he references French dictionaries such as Littré more than once and this would suggest that it is indeed the standard French of France which he is using as a model. He also occasionally describes Canadian-French speakers in a disparaging manner, for example, 'Le Canadien traduit trop littéralement et fait des néologismes impardonnables' (16 October 1884), and refers to usages specific to Canada being incorrect, which leads one to suppose that French elsewhere (for example, in France) is not.
On the evidence of these two language columns, then, it would appear that, at this period in time, there is a greater concern with adherence to prescriptive norms, and with a standard form of French, than with anglicisms. Although this standard is not explicitly named, it would appear that both authors are implicitly referring to the standard French of France.

1940-1949
A sample of articles from the language columns of two authors was taken for the period 1940-1949: Alfred Carrier and the Société du parler français au Canada (henceforth SPFC).

Samples were taken from two different SPFC columns 'Corrigeons-nous !' (B) and
'Corrigeons-nous !' (C) (column B was published as a series of leaflets, column C appeared in Le Canada français, see Appendix A for more detail). Again, there is a difference in the length and style of the two language columns, In general, the SPFC's articles are briefer and less detailed than Carrier's; the latter generally goes into great detail (3-4 pages) on one issue, while in the former, the majority of articles in column B either deal with two to three issues per article or provide lists of terms to avoid with their recommended replacements. The articles in column C are somewhat longer, but number only 8 in total. The SPFC deals with both general French 'errors' (terms or expressions that do not conform to prescriptive norms, frequently French-Canadian usages or archaisms) and anglicisms, while Carrier's column deals almost exclusively with perceived errors in French and only extremely rarely discusses anglicisms.

CARRIER Alfred
In his column, Carrier goes into some detail about various questions of usage and grammar, in particular discussing the 'correct' use of various terms, expressions and parts of speech (frequently discussing written style and points of spelling/abbreviation etc.). In these discussions, he invariably refers to prescriptive manuals and dictionaries and also to the French of France to reinforce the points that he is making, and he also not infrequently goes into the history of a term or phrase. For example, in his first article, Carrier deals with the use of the article with proper names, and during the course of his discussion, makes reference to three prescriptive works, including a grammar book produced by the Frères des Écoles Chrétiennes, 35 the Grammaire Larousse du XXe Siècle 36 and the second volume of Geoffrion's Zigzags autour de nos parlers 37 (an earlier language column) (24 September 1942). In an article discussing the correct abbreviation of the term comté (as in the regional county municipality, Le compté du Lac-St-Jean), he describes France as 'notre pays d'origine' and goes on to refer to prescriptive manuals to support his argument, including, again, a grammar book produced by the Frères des Écoles Chrétiennes (15 October 1942). 38 He returns frequently to this grammar in further articles. Other grammars referred to include the Grammaire de la langue française par une reunion de professeurs (F.E.C.); 39 the Grammaire de l'Académie; 40 the Grammaire française instantanée. 41 47 Of all the language columns examined, Carrier's deals with grammatical issues in the most detail. It also makes the most frequent mention of prescriptive guides. Such wide-ranging intertextual references may well allow him to increase his authority, but they also make his prescriptive intent very clear, as do comments such as the following: 'Nous ne devrions jamais nous défaire de nos livres de classe. Nous devrions principalement conserver la grammaire que nous avons apprise sur les bancs de l'école et en relire de temps à autre une page' (9 September 1943).
He refers, in this column, to the uncertainty of correct usage, due to the variation portrayed in various newspapers, journals and books in Quebec: 'nous sommes submergés par le flot de doute et de l'incertitude, ballottés d'une rive à l'autre sans que nous sachions à laquelle nous agripper' (29 October 1942). He compares this to the conformity found in France, and uses this as an argument in favour of using the French of France as the standard to follow: 'Comment expliquer que livres, journaux et revues, en France, s'accordent à suivre ainsi la même ligne de conduite ? D'où vient cette unanimité ? N'est-ce pas parce que c'est l'usage, le BON usage ?' (29 October 1942). He frequently refers to 'usage' being the arbiter of correctness, stating that 'Pour nous, les humbles et les profanes, nous n'avons qu'à nous incliner devant ce maître inexorable qu'est l'usage' (19 November 1942) and that 'c'est l'usage qui est le grand maître dans les questions de langage' (28 December 1944). However, his point of reference appears always to be usage in France, rather than that in Canada. He makes this clear when he says 'Dans les questions de langue, l'Usage est notre maître souverain à tous. Bien entendu, l'usage suivi en France quand il s'agit de la langue française.' The importance for him of following the norms of the French used in France is perhaps because he believes that 'ce n'est certes pas l'usage suivi de ce côté-ci de l'Atlantique qui va régir les destinées de la langue française' (4 May 1944).

Société du parler français au Canada
The sample of articles from the SPFC's language column is very clearly prescriptive in tone.
Indeed, the majority of the articles (30 out of 50) consist simply of two opposing columns of terms, one labelled 'Ne dites pas' and the other 'Mais dites'. The remaining twenty articles enter into a more detailed description of terms or expressions, but generally aim to encourage the use of one term while discouraging the use of another, usually the term used in France.
Indeed, it quickly becomes clear to the reader that the main aim of the column (during the period examined) is to align usage in French Canada with the standard French of France.
Prescriptive formulations are frequently used, for example, 'il faut employer X'; 'l'expression propre est X'; 'on devrait éviter d'employer X'; 'le bon usage prohibe X', 'il est incorrect de dire X'. Most of the more detailed articles in the sample make reference to the French used in France, nearly always in order to recommend the replacement of an anglicism or a French-Canadian term by the French one, using phrases such 'ne s'emploie guère en On the evidence of these two language columns, then, it would appear that at this period in time there is a continuation of the concern with the adherence to prescriptive norms (very explicitly standard French norms) and the avoidance of anglicisms displayed in the earlier period. However, the prescriptive norm to which readers are expected to adhere has now become very explicitly the standard French norm. This was not made explicit in the earlier sample (1880-89).

2008-2010
A sample of articles for the period 2008-2010 was taken for MEF and Impératif français. Impératif français produces a journal, Ensemble, 48 which is available free of charge on its website.. Although MEF does not produce a journal, its concerns are clearly represented in the numerous articles it publishes on its website. A sample of 100 articles was taken from Ensemble for the three-year period 2008-2010 and from MEF's website to discern the main areas of preoccupation of these societies. In addition, both societies devote parts of their website to information on anglicisms in French and the 'correct' use of French. These display many similarities to language columns, as their titles show: a section of MEF's website is devoted to Les mots qu'il faut 49 and Impératif français publishes a series of notes on language in its Chronique linguistique. 50 These sections were also analysed.
Analysis of the journal articles showed that the articles produced by the two societies mainly deal with the issue of the marginalization of the French language in Quebec due to the widespread use of English. This fear is discussed generally, rather than manifesting itself in a preoccupation with anglicisms, which are discussed in only a minority of articles. For example, many articles suggest reasons for this marginalisation, such as, for example, the supposed collusion of political and industrial 'elites' who are seen to actively encourage anglicization or the continued weakening of Bill 101 due to its lack of enforcement (Walsh 2013: 136-38). Both MEF's Les mots qu'il faut and Imperatif francais' Chronique linguistique are made up of lists of anglicisms to be avoided and incorrect French or errors to be avoided. In its discussions of 'errors', MEF concentrates largely on the correct use of certain French terms (e.g. the difference between valable and valide; the correct use of the terms civique and civil) and does not touch upon issues of spelling or grammar at all. Impératif français discusses capitalization and points of punctuation and spelling but also concentrates heavily on the correct use of particular French terms. 51 In terms of the metalanguage used by these societies, both tend towards a subjective and emotive use of language, where a strong correlation between language and national identity is made. For example, MEF claims that 'Langue française et identité québécoise sont inséparables. Sauvegarder l'une c'est sauvegarder l'autre' 52 and that 'Une langue est d'abord une patrie. C'est en elle et par elle que se soudent la solidarité nationale, le vouloir-vivre ensemble et la communauté de destin.' 53 Impératif français asserts that 'la langue française […] est le signe le plus significatif de notre identité nationale'. 54 Both societies also occasionally use images of war and colonization, imagery which is commonly associated with purist discourses. MEF refers to 'La nouvelle Gestapo intellectuelle qui interdit l'usage des langues autres que l'anglais dans la communication scientifique internationale', 55 and Impératif français compares the battle to promote the French language to a war needing an army: pouvez-vous nous assurer, que celles et ceux qui devront toujours lutter pour la promotion et la protection de la langue de Molière, à Montréal, pourront compter sur un chef d'Armée, pour ce faire ? On n'a rarement vu une armée triompher, lorsqu'il n'y avait pas de Général pour diriger les troupes. 56

Thomas's theoretical framework
Having examined the contents of each of the four columns and the three language societies, the theoretical framework developed by Thomas will now be used to assess the types and level of purism present in these. Using the same framework as Walsh (2013) 57 Figure 1. 58 In addition to the questionnaire, a numerical index which allows a quantitative measure of the intensity of the purism that exists in a given community is reproduced in Figure 2. The numerical index allows us to measure the weighting of non-puristic factors (this corresponds to question one in the questionnaire), the configuration of puristic orientation (this corresponds to questions two to nine), and the targets of purism (questions ten and eleven), and preferred replacements (question twelve). As bench-marks for measuring the puristic profile of a situation, Thomas uses the terms mild, moderate and extreme and, depending on what is being measured, he outlines the criteria for deciding which bench-mark to use. Each bench-mark is also assigned a value on his scale (mild purism is assigned 1-2 on the scale, moderate 3-4 and extreme 5). 59 Where there are two values, non-borderline cases are given the lower value (1 for mild purism, 3 for moderate purism) and borderline cases are given the higher value (borderline mild-moderate 2, borderline moderate-extreme 4).

Purist and non-purist concerns
The first question asks to what extent the purism takes non-puristic concerns into consideration in developing its ideology. Thomas argues that non-puristic concerns are motivated by instrumental (i.e. practical) and ethical attitudes, and puristic concerns are motivated by affective, traditionalist and nationalist attitudes. All linguistic behaviour can be seen to be motivated in some way by one or more of these attitudes. 60 With regard to purist and non-purist concerns, Thomas states that mild purism assigns equal importance to purist and non-purist concerns; moderate purism makes some concessions to non-purist concerns; and extreme purism does not take non-purist concerns into account at all. 61  The SPFC also displays both an affective and a traditionalist attitude, whereas Carrier displays a very strong traditionalist attitude only. We might also argue that Carrier displays a somewhat nationalist attitude, given his very clear belief that only the French language of France can adequately serve as the national language of Quebec, and his reference to France as 'notre pays d'origine'. However, this will not change the outcome of the analysis, as this is also a purist consideration. None of the columns show any perceptible non-purist concerns, for example, that avoiding the use of a particular term will aid comprehension. It may be that the authors feel that using only those terms which are in accepted use in France will aid comprehension between Canadian-French and French speakers; if this is the case, none of them state so explicitly and we cannot take this for granted. According to Thomas's model, therefore, all of the columns examined display extreme purism and can be assigned a 5 on Thomas's index.
The two language societies examined also all display an affective attitude, as both oppose the use of English. They also make extremely strong links between the French language and Québécois identity, thereby displaying a nationalist attitude (although one that is very different to that shown by Carrier, whose national sentiment is for France rather than Quebec). They therefore base their ideology on at least one purist concern. Although neither society can really be argued to display any instrumental concerns, the ethical concern to allow the integration of the whole speech community is clearly shared by both. It is one of their main aims in ensuring the promotion, as well as the protection, of the French language.
They therefore base their ideology on at least one non-puristic concern alongside two puristic concerns and can be seen to display moderate purism (3 on Thomas's index).

Orientation of purism
Questions two to nine concern the orientation of the purism. Thomas makes a general distinction between 'external' purism and 'internal' purism. 'External' (or 'xenophobic') purism involves 'removing or replacing foreign elements, whether their source is specified (targeted xenophobia) or unspecified (general xenophobia)'. 65 'Internal' purism includes archaising and reformist purism, which have a temporal basis, and ethnographic and elitist purism, which have a social basis. Archaising purism typically involves 'an attempt to resuscitate the linguistic material of a past golden age, an exaggerated respect for past literary models, and excessive conservatism towards innovations', 66 whereas reformist purism makes 'a conscious effort to reform, regenerate, renew or resuscitate a language' and is important in the creation of standard languages (ibid.). Ethnographic purism favours rural dialects as a source of new words, as these are seen as being in some way purer than the standard language. 67 Elitist purism, on the other hand, is negative towards these rural dialects or any non-standard speech, and is based on the belief that language can be perfected. 68

Targets and replacements of purism
Question ten refers to the linguistic level on which the purism takes place. Thomas claims that examples of purism can be found on all linguistic levels (phonological, morphological, syntactic, lexical and orthographic), but makes a distinction between unmarked purism, operating at the lexical level, and marked purism, operating at the levels of smaller or larger linguistic segments; he thus views lexical purism as archetypal. 71 In Walsh 2013, this view is strongly reinforced. 72 It is certainly the case that the majority of targets in all of the language columns are lexical; although they also deal occasionally with issues in orthography, pronunciation (phonology) and occasionally word formation (morphology) and sentence formation (syntax , e.g. negation), the vast majority of articles deal with individual words.
This is also the case for the language societies.
Question twelve concerns the types of word which are targeted by the external lexical purism and question thirteen the replacements of these targets. Thomas's model has been adapted here, as the original lacked clarity. 73 In the adapted framework, in terms of the targets of lexical purism, mild purism is directed at loanwords from unrelated languages; moderate purism at loanwords from related languages, and internationalisms; and extreme purism at calques (loan translations, loan renditions and semantic loans). In terms of the replacements, mild purism is characterised by the acceptance as replacements of internationalisms, unassimilated loanwords from related and unrelated languages and assimilated loanwords from unrelated languages. Moderate purism is characterised by the rejection of these, but the acceptance of assimilated loanwords from related languages, loan translations, loan renditions, semantic loans and loan creations. Extreme purism is characterised by the acceptance as replacements either existing native language terms or loan renditions and loan creations in the absence of other types of replacement, which are rejected.
That is, while moderate purism accepts other types of calque and assimilated loanwords in addition to loan renditions and loan creations, extreme purism accepts loan renditions and loan creations only and rejects any other type of calque or loanword.
In each sample of articles examined, the anglicisms discussed and the replacements suggested have been analysed and categorized according to the typology outlined in Walsh 2013. 74 Figure 3 shows the proportion of different categories targeted in the language columns. It is worth noting that Carrier's column is excluded here because targets and replacements refer to external purism only. Anonymous mainly targets calques in his column: loan translations and semantic loans make up 95% of targets. He therefore displays extreme purism here and can be assigned a 5 on Thomas's index. The majority of Lusignan's targets are also calques: loan translations and semantic loans account for 61% of all targets.
However, assimilated and unassimilated loanwords make up nearly two-fifths of the total (39%). Although, according to Thomas (Walsh 2013: 133-34). For this reason, we can classify them as borderline cases and give them a value of 4 on the numerical index. Figure 5 shows the replacements suggested for the anglicisms targeted by the language columns. In all three cases, the majority of replacements are existing French terms.

Figure 4 Language society targets
Accepting existing French terms as replacements only happens in cases where an anglicism is competing for use with an existing French term with the same meaning. We can assume that allowing both terms to coexist as synonyms would indicate a mildly purist attitude. Insisting upon the complete removal of the anglicism and the exclusive use of the French term, however, reflects a non-compromising position which must be seen as extreme purism. Both Anonymous and APFQ can be assigned a 5 on Thomas's index. Although Lusignan does accept some assimilated loanwords as replacements, by far the majority are existing French terms and it is therefore reasonable to assign him a 5 also.  Figure 6 shows the replacements suggested by the language societies. While MEF mainly chooses existing French terms as replacements (90%), Impératif français also suggests some calques (12%) and loan creations (18%). MEF clearly displays extreme purism here (5 on Thomas's index), whereas Impératif français can be seen to represent a borderline case between moderate and extreme purism (4 on Thomas's index).

Figure 6 Language society replacements
Presenting the values for all of the language columns and the language societies sideby-side in Figure 7, we can see that Anonymous and Lusignan's columns are the most purist,  Using the numerical index as a guide, it is difficult to note any major differences in the levels of purism over the three periods examined. However, as we shall see in the conclusions below, the analysis has highlighted some differences between the three periods.
This raises an issue with the model. The average values displayed by the numerical index for the language societies and the language columns for the period 1949-1949 are identical, and yet the purism displayed by two show some clear differences, as we have seen. This suggests that relying on the index alone to provide a picture of purism in a given situation may cause a misleading assessment of that situation. However, the numerical index does provide a useful support to a qualitative analysis of purism. It is not intended to be looked at in isolation from this analysis, but rather to act as an additional means of comparison.

Conclusion
Both the earlier and later samples of articles from language columns examined display a clear preoccupation with adherence to prescriptive norms and the influence of English on the French language in Quebec, as do the modern day language societies. However, the analysis has highlighted some differences between the three periods, and in particular, between the two earlier periods 1880-1889 and 1940-1949 and the modern-day period. This may reflect a change in attitudes over time.
Firstly, the results show that the earlier language columns (1880-1889) are more purist overall, with the later language columns (1940)(1941)(1942)(1943)(1944)(1945)(1946)(1947)(1948)(1949) and language societies displaying a more moderate level of purism. The earlier language columns are particularly purist in the type of anglicisms that they target. Loan translations and semantic loans (calques) are much less noticeable than assimilated or unassimilated loanwords. Indeed, Thomas argues that calques may be tolerated where loanwords from the same language are not, perhaps because calques, even those consciously created to replace loanwords, are not clearly foreign, and can pass unnoticed by many (indeed, perhaps even most) speakers. 75 However, he also notes that some purists may see calques as being more dangerous even than loanwords, as their assimilation into the native system may cause a threat to the system in some way. 76 For this reason, Thomas sees them as a useful test of the level of purism: 'Indeed, precisely because of their compromising nature, calques offer an excellent litmus test of the intensity of puristic feeling'. 77 It would appear from this that puristic feeling was more intense in 1880-1889. The fact that both sets of language columns display more purist concerns in their ideology than the language societies, of which two also display non-purist concerns and, specifically, a desire to ensure the integration of the whole speech community, may also suggest a higher level of purism in earlier times (both 1880-1889 and 1940-1949).
A second, more striking, change between the earlier periods and the modern period is the difference in emphasis laid on the French of France. A concern for Québécois French to adhere to the prescriptive norms of the standard French of France is rarely, if ever, mentioned by the language societies. There could be a number reasons for this. It may be, for example, that the introduction of Bill 101 and the subsequent transformation of society in Quebec and increased status of French has led to a reduction in the linguistic insecurity felt by speakers of Canadian-French, at least insecurity relating to adherence to an external prescriptive norm. language columns also all make a higher use of prescriptive language than the language societies and make far more frequent reference to dictionaries, usage guides and grammers.
Although the language societies include some lists of the type 'ne dites pas ... mais dites', the majority of articles in their journals and on their websites do not employ a very highly prescriptive tone, nor do they refer to dictionaries or grammars. The language columns, on the other hand, makes it very clear to the reader that one variety of French is acceptable and one variety only, the French of France; although they do not directly use purist themes of corruption or contamination, there is no doubt that they feel that this language has become corrupted by local French-canadian terms or anglicisms and that it is necessary to remove and replace the corrupted parts. There is far less concern with prescriptive norms in the modernday societies and no apparent preoccupation with the French of France. It could be argued, therefore, that the introduction of Bill 101 and the subsequent transformation of society in Quebec and increased status of French has led to a reduction in the linguistic insecurity felt by speakers of Canadian-French, an insecurity which is so clearly portrayed by the concern to avoid regionalism or archaisms shown by language columnists in the periods 1880-1889 and 1940-1949 and yet far less evident in the modern-day language societies. Now that French has become a majority language in Quebec, used in the workplace as well as in the home, it may be that the need to adhere to an outside 'prestige' standard is diminished.
On the other hand, this change may equally be due to the fact that, in spite of the transformative changes brought about by Bill 101, for modern-day language societies in Quebec the issue of the influence of English on French has become ever more pressing and urgent. It is possible that the threat to the continued existence of the French language in Quebec is felt by the language societies to be so great that they see little point in concerning themselves with the quality of the language. A concern with prescriptive norms is likely to remain a minor consideration, as long as speakers feel threatened by a more powerful language, namely English. Although this threat already existed in the late nineteenth century and first half of the twentieth century, the increasing influence and power of the United States since the second World War in global economic, political and cultural spheres and the increase in the use of English as an international language led to a corresponding decline in the use of French, which has arguably increased the feeling of threat from the English language in Quebec. In the current day, the prevalence of English on the internet and the fact that it is the working language of most international companies has no doubt exacerbated this fear, in spite of the far-reaching and transforming changes effected by Bill 101.
In summary, this study has shown that linguistic purism in French Canada has changed over the years, moving from a preoccupation with internal 'correctness' and adherence to the standard French of France and a minority interest in anglicisms, to a much reduced concern with the internal structure of the language or adherence to any external standard and a much greater concern with the influence of English on French. This may suggest that the implementation of linguistic legislation such as the far-reaching Bill 101 in Quebec has led to a decrease in the linguistic insecurity felt by French-Canadian speakers or at least a move away from insecurity based on lack of adherence to an external norm.
Equally, the increased threat from the English-speaking world in political social and cultural arenas as well as the linguistic one may simply have concentrated the focus of those concerned with defending the language onto this threat, thereby reducing the amount of time and space they give to internal prescription. A much broader study examining a wider range of language columns over more periods in time is now necessary to allow us to determine more clearly the reasons for such a change.