ajs bilingualism in montreal a demographic analysis

17
7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 1/17 Bilingualism in Montreal: A Demographic Analysis Author(s): Stanley Lieberson Source: American Journal of Sociology, Vol. 71, No. 1 (Jul., 1965), pp. 10-25 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/2774765 . Accessed: 23/06/2013 16:23 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to  American Journal of Sociology. http://www.jstor.org This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PM All use subject to JSTOR Terms and Conditions

Upload: marcelo-gonzalez-figueroa

Post on 03-Apr-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 1/17

Bilingualism in Montreal: A Demographic AnalysisAuthor(s): Stanley LiebersonSource: American Journal of Sociology, Vol. 71, No. 1 (Jul., 1965), pp. 10-25Published by: The University of Chicago Press

Stable URL: http://www.jstor.org/stable/2774765 .

Accessed: 23/06/2013 16:23

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of 

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to

 American Journal of Sociology.

http://www.jstor.org

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 2: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 2/17

Bilingualismn Montreal:A Demographic Analysis'

Stanley Lieberson

ABSTRACT

Census data and indexes adopted from linguistics are used to examine trends in the ability of Mont-real's population to communicate with one another between 1921 and 1961. There has been no increase inlinguistic communication during this period. The probabilities of members sharing a mutually intelligibletongue within the British and French ethnic groups remain far higher than the probability of linguisticcommunication between the groups. Bilingualism appears to be an end product of language contact ratherthan an intermediate step toward monolingualism. Both major languages maintain their positions in in-tergenerational transfer. Cohort analysis discloses relatively low degrees of bilingualism among small chil-

dren, a rapid rise during the school and young adult ages, and an actual net decline during the middle andolder ages.

Montreal might be viewed as a battle-

ground between the French language and

cultureof Quebecand the English-speaking

Canadians and Americans who surround

French Canada. This metropolitan area,

containing more than 10 per cent of

Canada's population, is incontestably the

great center of English-French contact in

North America. Mforethan 800,000 speak

only the French official language, nearlya half-millionspeak only English, and more

than three-quartersof a million are bilin-

gual (speaking botlhEnglish and French).

The latter group comprises slightly more

than a third of the nation's bilingualpopu-

lation.2

Perhaps "battleground"is too dramatic

a term for describing French-Englishrela-

tions in Montreal, although the occasional

acts of violence, the more frequent verbalexpressions of nationalism, and the self-

consciousness about language make our

metaphorapt. If inherent in linguistic con-

tact is the danger or possibility that one

languagewill decline and the other expand,

then, in this fundamental sense, Montreal,

or any other multilingual setting, is abattleground.

Although we are not prepared to assertthat the loss of French in Montreal wouldmean a decline in the importance of thislanguage elsewhere in the province, it isreasonableto assume that Montreal facili-tates the maintenance of the Frenchtongueand the bicultural society in Canada. Since

40 per cent of Quebec's population is lo-cated in this metropolitan area, Montrealis of intrinsic interest because of its sheer

size. Moreover, if there is any validity tothe concept of metropolitandominance,wewould expect the linguistic outcome in

Montrealto influence thosepartsof Quebecin its hinterland. As Hughes put it, "Mont-real is the port of entry from which Enig-lish influence and the industrial revolution

radiate into the remote corners of theFrench-Canadianworld."3 The presenceof

Montreal as a French-speakingcenter pro-vides Canada with a complete French-

language society, which runs the gamutfrom backward rural areas to a modern

metropolis maintaining relations with the

rest of the world. The existence of this

metropolisallows for a more self-containedThis study was supported by the National Sci-

ence Foundation Grant No. G-23923. The assistance

of Elaine F. Green and David Sorenson is grate-

fully acknowledged.

'All figures based on 1961 census data.

3Everett C. Hughes, French Canada in Transi-

tion (Chicago: University of Chicago Press, 1943),

p. 202.

10

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 3: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 3/17

BILINGUALISMN MONTREAL 11

French community than would otherwiseoccur. Of considerable importance, too, isthe extensive out-migration from rural

FrenchCanadacurrentlyunderway.4 Mont-

real provides an outlet without necessarilya loss of these people from the Frenchlanguageof the ethnic community.

In this paperwe first examine the trendsin language usage over the past forty yearsand then interpret them from a primarilydemographicperspective.Our interest stemsfrom the fact that the encounterof different

speech groups is often a concomitant ofracial and ethnic contact. The linguistic

outcome in a society composed of peoples

with different tongues is by no means aforegone conclusion in Canada or in such

varied parts of the world as Belgium,Soviet Moslem Asia, India, South Africa,Ceylon, and Paraguay.5Although recogniz-ing that many theoretical approaches are

appropriate, we are concerned here with

those demographicfactors that maintain a

multilingualsystem and those that lead to

monolingualism.

BILINGUALISM IN DEMOGRAPHIC

PERSPECTIVE

If, by definition, no society can exist

unless its membersare able to communicate

with one another, then the developmentof

a bilingual population can be expected in

any settlement in which different languagegroups reside. In this sense, bilingualism

may be viewed as an adaptation by in-

dividuals to the presence of persons with

another tongue. Although bilingualismis

expected, there is no assurance that it will

be of the reciprocalvariety in which each

grouplearns a second language to an equal

degree. Clearly, far more non-English-

speakers within the United States learnEnglish than native Americans learn theimmigrant tongues.6 Moreover, the exist-

ence of bilingualism raises a fundamental

question for Montreal or for any similarsetting, namely, "Does second-language-learning occur without a permanentchangein the linguistic compositionof the popula-tion, or does it lead to a loss in the nextgeneration in the position of one of themother-tonguegroups?"

Bilingualism may be viewed as havingthree possible outcomes. First, it can pro-vide the mechanismleading to the develop-ment of a monolingual population. This is

what occurredin the United States amongnon-English-speakinggroupswho migratedafter the establishment of Anglo-Saxondominance.After these groups learned thenation's language and became bilingual,

somewhere along the line of linguistictransfer between generations only English

was passed on and not the mother tongueof the old world. Although completion of

this processmay have taken several genera-tions, clearly bilingualism was an inter-

mediate stage between the arrival of non-English-speakersand the final loss of the

foreign tongue.Second, bilingualism may be an end

product in itself. In this case, there is

sufficient bilingualism to enable a popula-tion with differentmother tongues to main-tain the social system, but the bilingualismof parents does not lead to the loss of the

mother tongue among the next generation.

Rather,the next generation receives the

parents' first language and merely repeatsthe process.Under such circumstances,two

populations may be in contact indefinitelywithout the decline of either language.

The thirdpossibility,whichneed not con-

cern us here but will be mentionedmerelyfor the sake of completeness, is that the

speakersof one language may begin to use

Nathan Keyfitz, "L'Exode rural dans la pro-

vince de Qu6bec, 1951-1961," Recherches socio-

graphiques, III (September-December, 1962), 303-

15.

'See, for example, United Nations, Demographic

Yearbook, 1956 (New York: Department of Eco-

nomic and Social Affairs, 1956), Table 9; Demo-graphic Yearbook, 1963 (New York: Department

of Economic and Social Affairs, 1964), Table 10.

8 In 1930, less than 10 per cent of immigrants 10

years old and over were unable to speak English.

U.S. Bureau of the Census, Fifteenth Census ofthe United States, 1930, Population (Washington,

D.C.: Government Printing Office, 1933), II, 1347.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 4: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 4/17

12 THEAMERICANOURNALOF SOCIOLOGY

a simplified form of the other group'stongue (pidginization), which in turn is

passedon to children as their first or mothertongue (creolization).7 Since this is un-

likely in a society such as Canada'swhereboth languages are standardized, we shallrestrict our attention to the first two possi-

bilities. This is not to deny, of course, theincorporation into each language of loan

words from the other.8

DATA AND METHODS

Unless otherwise indicated, the decennial

censusesof Canadaare the sources for thedata used in this study. Two basic language

questionshave been asked for a numberofdecades.The first concernsthe population'smother tongue. The census data on mother

tongue can be used as a reasonably good

measure of the first language learned.There is one difficulty,however, caused bythe stipulation that, to qualify as one'smother tongue, the first language learnedmust still be understood (beginning with

the 1941 census) or spoken (1931 and

earlier). Since we are dealingwith a settingin which the number of both English and

French speakers is large, it would seem

reasonableto assumethat nearly all personsin Montreal whose first language was eitherFrench or English will retain sufficient

knowledgeof the languageso that they will

meet this mothertongue criteriaof the cen-

sus takers if their first languagewas eitherof these. Very likely part of the populationwho learned some other language first will

no longer be able to speak or understanditand, therefore, will have some other lan-

guage which they currently know reported

as theirmothertongue.This mightespecially

be the case for foreign-bornadults who hadcome to Canada as small children. When

using data from Canadian censuses, then,equating "mother tongue" and the first

language learned is open to error. How-ever, we shall see that this erroris counterto the results reported and does not biasthe data in favor of the interpretationsoffered below.

The second-languagequestion relates tothe respondents'ability to speak the officiallanguages of Canada-English and French

-at the time of the census.The resultsleadto four categories: persons able to speak

only English; only French; both English

and French; neither official language. Thereader should note that the data on current

linguistic ability refer only to the twoofficial languages. Hence a person recorded

as speaking "English only" may be quitefluent in, say, Germanand Italian.

The difficulty n the official-language atalies in the great simplicity of the censusquestion, namely, "Can you speak Eng-lish? French?" We are unable to determinethe subjective criteria a respondent useswhen he reports his ability to speak theselanguages. Moreover, there is some un-

certainty about the biases of respondents.On the one hand, there is social-psycho-logical evidence to indicate that English

has greater prestige than French in Mont-

real.9 By contrast, the recent intensifica-

tion of French-Canadiannationalism and

separatism in Canada would lead us to

expect an underreportingof English-speak-

ing ability among the French population.The frequent exposure to the languagesthrough acquaintances, radio and televi-

sion, signs, newspapers, and conversations

of strangersin public places would lead us

to expect that the questions are less am-

biguous for Montreal than for other parts

of Canada,because Montrealresidentscon-

stantly encountera "realitycheck."Finally,

7See, for example,WilliamJ. Samarin,"LinguaFrancas, with Special Reference to Africa," in

Frank A. Rice (ed.), Study of the Role of Second

Languages Washington,D.C.: Centerfor AppliedLinguistics,1962), pp. 54-64.

8 For a discussionof English loan words used

by French Canadians n an outlying rural area,

see HoraceMiner,St. Denis: A French-CanadianParish (Chicago: University of Chicago Press,

1939), pp. 32-35, 242-45.

'Wallace E. Lambert, R. C. Hodgson, R. C.

Gardner,and S. Fillenbaum,"EvaluationalReac-tions to SpokenLanguages," ournalof Abnormaland Social Psychology,LX (1960), 44-51.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 5: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 5/17

BILINGUALISMN MONTREAL 13

adopting the standard demographicchecks

for census data, we have tested the mother-

tongue and the official-languagedata for

Canadaforinternalandexternalconsistency

without revealing any serious errors.10As is customarywith analyses over time,

various adjustments, splices, and estimates

to maximize comparability through the

decadeshave been used. This is particularlyimportantwith respect to the spatial area

covered.As much of the metropolitanarea

of Montreal as was defined in 1961 has

been included for earlier periods as well.

However,unlike the StandardMetropolitan

Statistical Areas in the U.S. Census, Cana-

dian metropolitanareas are not defined interms of total counties. Hence, compari-

sons of the total metropolitan area would

be based on dissimilar spatial units which

cannot be reconciled for the differentdec-

ades.The linguist Greenberghas proposed a

series of measures of linguistic diversity

that can be used to determinethe possibili-

ties of communicationamong the popula-

tion of some delimited area.1'One of these

measures,the index of communication(H),has been adapted to the Montreal study.

The H gives the probability of two ran-

domly drawn residents of Montreal being

able to speak to each other in one or both

official anguages.Thus, an H of 1.00 would

mean that all persons in the area share a

mutually intelligible language.At the other

extreme, we would obtain H = 0 when

there was no chanceof two randomlydrawn

personsin an area sharing the same officiallanguage. More formally, let a = the pro-

portion of the population speaking only

English; b = the proportionof the popula-

tion speaking only French; c = the pro-

portionof the population that is bilingual;

d = the proportionunable to speak either

official language. Then H. = a2 + b2+ c2

+ 2ac+2bc.An extension of Greenberg'sindex has

been proposedthat allows us to measurethe

probability of communication betweenmembers of different ethnic segments ofthe metropolitan area.12 Using the same

scale we can determine the probability ofmutually intelligible communication be-

tween randomly selected members of twogroups as contrasted with Greenberg'smeasuresof communicationwithin a group.

These measures, used in tandem, allow usto determine the extent of linguistic cohe-sion within populations such as the total

city or ethnic group as well as the linguisticpotential between such populations. Weshall use H, to indicate Greenberg'smeas-ure of communication among randomlyselected membersof a single populationand

Hb to indicate communication betweenrandomly selected members of two popula-

tions. For groupsi and j,

Hb= ai(aj + cj) + bi (bj + cj)

+ ci(a?+bj+ cj).

TRENDS

Communication.-The H, gives theprobability of two randomly drawn resi-

dents of Montreal being able to speak to

each other in one or both official lan-

guages of Canada. For 1961, we would ex-

pect two persons randomly selected fromthe metropolitanarea to speak in a mutu-

ally intelligibleofficial anguage slightly less

than eight out of ten times. This is a fairlyhigh degree of mutual intelligibility; mul-

tilingual Brussels (Brabant), for example,has a slightly lowerH for 1947.13 However,we should not lose sight of the fact that

somewhat more than 20 per cent of the

time there is no shared communication

possible in these languages. More striking,and contrary to the expectationsgenerated

10 Stanley Lieberson, "Language Questions in

Censuses" (report prepared for S.S.R.C. Socio-

linguistics Seminar, Indiana University, Summer,

1964).

'

Joseph H. Greenberg, "The Measurement ofLinguistic Diversity," Language, XXXII (Jan-

uary-March, 1956), 109-15.

Stanley Lieberson, "An Extension of Green-

berg's Measures of Linguistic Diversity," Lan-

guage, XL (October-December, 1964), 526-31.

"Ibid., p. 528.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 6: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 6/17

14 THEAMERICANOURNALOF SOCIOLOGY

from the American urban experience, wefind something of a decline in HWIn thedecades between 1921 and 1961 (Table 1,col. [1]).

One important modification of this re-sult must be considered, namely, changesin II, may reflect fluctuationsin immigra-tion rather than a basic alteration of thelinguistic processes of Montreal. Specifi-cally, the component of the populationunable to speak either official languagerises in recent decades, and much of the

(Table 1, col. [2]). Since no discernibletrend is found after these adjustments, wecan conclude that the decline is due toimmigration. At best, however, we do not

find a substantial rise in Hw over time, al-though communication potential has moreor less remainedconstant in the forty yearsbetween 1921 and 1961.

Using the extension of Greenberg'smeasures described earlier, we can measurethe probability of communication betweenmembers of the different ethnic segmentsof

drop in communication may reflect thisrise. In 1961 immigrants arriving duringthe precedingfive and one-half years com-

prised 5 per cent of the area's populationbut accounted for one-half of the popula-tion of the entire metropolitanarea unableto speak English or French.

The indexes have been recomputed intwo different ways: first, excluding allpersons unable to speak either official lan-

guage; second, excluding the foreign-bornpopulation entirely. In either case, no ap-preciablechange in the degree of linguistic

communicationis observed over time. In

some comparisonsthere is a slight rise andin others a slight decline through time

Montreal. We observe a rise between 1921and 1931 in the chance of British-Frenchinteraction in a mutually intelligible lan-

guage (Table 1, col. [31). This is not thecase in more recent decades-witness theslight decline from 0.6362 in 1941 to 0.6245in 1961. Moreover, for the entire metro-politan area in 1961, in only six out of tencases could we expect two people to sharea commonofficial language. On the basis ofthe data available, there is no reason tothink that the communicationpotential be-

tween British and French groups in Mont-real has been rising.

The British have much closer linguisticlinks with other ethnic groups in Montreal

TABLE 1

INDEXES OF COMMUNICATION (H) BETWEEN AND WITHIN SELECTED

POPULATIONS OF THE MONTREAL AREA, 1921-61

BETWEENETHNICGROUPS

WITHIN

WITHIN TOTAL POPULATION British FrenchYEAR SPEAKINGAN

POPULATION OFFICIAL

LANGUAGE French Other Other

(1) (2) (3) (4) (5)

1921*....... 0.8719 0.8833 0.7552 0.8879 0.74441931*....... .8752 .8918 .7665 .9042 .75711941* ... ... .8598 .8656 n.a.t n.a. n.a.

1961*....... .8349 .8727 n.a. n.a. n.a.

1941t ....... .8294 .8361 .6362 .9278 .68301951t....... .8222 .8385 .6269 .8849 .66641961t ....... 0.7942 0.8425 0.6245 0.7795 0.6174

* Montreal and Verdun, ten years of age and older.

t Not available.: Montreal, Verdun, and Outremont, all ages.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 7: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 7/17

BILINGUALISMN MONTREAL 15

than do the French (cf. Table 1, cols. [4]and [5]). The closer linguistic communica-tion of other groups in general with theBritish is a result of the greater affinity of

these groups for learning English than forlearning French. We shall consider this

matter in greater detail later. Because ofextensive immigration,we are not inclinedto take too seriously the drop in Hb for

"other" ethnics between 1941 and 1961.These results indicate that linguistic co-

hesion between the major ethnic popula-

tions is only moderate, although cohesionwithin the British and French groups isextremely high. For the metropolitanarea

in 1961, communication potential withinthe British and the French ethnic groupsis 0.97 and 0.98, respectively. In short,language is a major factor in maintaining

ethnic cleavagein the city. We should note,however, that the probability of communi-

cation between the French and the Britishethnic groupsin the Montreal metropolitan

area is still nearly three times greater thanHb between Algeria's Europeansand Mos-lems in the period preceding the Algerianrevolution.14Other ethnic groups in Mont-

real have mainly veered toward the Eng-lish language and the British ethnicpopulation. Their linguistic unity is not

sufficiently high to maintain a residualthird ethnic force in Montreal. Linguistic

unity for the entire Montreal area appearsto be static after we take into account the

influence of immigration, which tends to

depress the H indexes because of the num-

bers unable to speak either official lan-guage.

Official language distribution.-H in-

dexes, being summary measures, may fail

to disclose important trends in language

usage because different combinations of

official-language distributions may yieldthe same index values. Examinationof the

basic data indicates that there has been a

drop in both the monolingualEnglish and

bilingual percentages, accompanied by a

rise in the segment speaking French only.

For example, Montreal-Verdun-Outremontdropped from 24 to 19 per cent monolingualEnglish and from 42 to 39 per cent bilingualin the twenty years between 1941 and 1961.

At the same time, the percentage speakingFrench only rose from 34 to 39 (Table 2,Panel A). These trends hold regardless ofthe areal basis upon which the comparisonsare made.

There are striking differences betweenethnic groups in their official-languagedis-tributions. The French are far morebilingual than are the British (cf. Table 2,Panels B and C). Through time it appearsthat the British population has at least

maintained its degree of bilingualism,whereas bilingualism has declined amongthe French. We should note that the Frenchnevertheless remain more bilingual than theBritish in 1961.

Other ethnic groups in Montreal occupyan intermediate position between the twoleading ethnic populations, but they defi-nitely favor English. They are more likelyto speak French than are the British, al-though an overwhelming segment speakEnglish only rather than French only (Table2, Panel D). Similar to the strengtheningof the French language among the Frenchethnic population observed earlier is asharp decline since 1941 among otherethnics in the percentage speaking Englishonly and a doubling in the small percent-age speaking French only.

It is difficult to account for these trends

in a quantifiablefashion since the necessary

cross-tabulations are not available. Weshould recall, however, that there has beena growingmovement from rural Quebec to

urban Quebec-particularly to Montreal.

The fact that ruralFrench Canadians ivingin more homogeneoussettlements are less

likely to learnEnglish than are their Mont-

real compatriots would help explain the

drop in bilingualismand the rise in mono-

lingual French-speakersamong the French

Canadians in Montreal. In 1961, for ex-

ample, only 9.7 per cent of the Frenchethnic population in rural Quebec was4Ibid., p. 528.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 8: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 8/17

16 THEAMERICANOURNALOF SOCIOLOGY

bilingual. We must note that internal mi-gration probably does not fully account for

the decline in French bilingualism, since

we have also observed a recent increase

among immigrants in their propensity tolearn French. It would appear that the

French Canadians, although still subordi-nate to the British in the economic world,have risen somewhat, and this may make

companies listed under "DepartmentStores" and forty-nine under "Magasins aRayons."'15To be sure, the number listedas true department stores must be taken

with a grain of salt; however, it is clearfrom inspection that all of the major down-town stores are now in both the Frenchand the English listings.

The possibility of distortions in the

knowledge of French more useful than it

has been for Montrealers.We are told, forexample,of a departmentstore that at onetime made no attempt to hire clerks who

knew French on the grounds that the

French Canadianscould not afford to shopin its high-pricedstore.The bilingualMont-

real classified telephone directories are

relevant here. Under the English-language

category, "Department Stores," there are

twenty-eight different listings in 1938.

Under the French equivalent, "Magasins

a Rayons," we find only four listings. Inthe 1964 Yellow Pages, we find forty-five

census data due to French nationalism

should be consideredas accounting for thedecline in bilingualism among French

Canadians.We are inclined to minimize this

explanation for several reasons. First, this

decline was observed before the recent in-

crease in nationalism. Second, the most

recent censuswas conductedbefore French-

TABLE 2

PERCENTAGE DISTRIBUTION OF OFFICIAL LANGUAGE BY ETHNIC ORIGIN

MONTREAL AREA, 1921-61

MONTREAL AND VERDUN, MONTREAL, VERDUN, AND

POPULATION AND 10 YEARS OF AGE AND OLDER OUTREMONT, ALL AGES

OFFICIAL LANGUAGE

1921 1931 1941 1961 1941 1951 1961

A. All groups:English only .......... 25.2 25.9 23.7 18.7 23.9 22.1 19.0French only ........... 21.9 20.5 28.2 32.5 34.0 35.8 39.0English and French ... 51.9 52.7 47.8 46.6 41.7 41.1 39.1

Neither ............... 0.9 0.9 0.3 2.2 0.4 1.0 2.9B. British:

English only .......... 71.4 71.4 n.a.* n.a. 70.2 70.6 67.1French only ........... 0.7 0.5 n.a. n.a. 1.3 2.0 2.8English and French ... 27.9 28.1 n.a. n.a. 28.5 27.4 29.9

Neither ............... 0.0 0.0 n.a. n.a. 0.0 0.0 0.2

C. French:English only .......... 0.2 0.5 n.a. n.a. 0.7 1.1 1.0French only ........... 34.3 32.7 n.a. n.a. 51.7 52.8 55.4

English and French .... 65.6 66.8 n.a. n.a. 47.6 46.1 43.4

Neither ............... 0.0 0.0 n.a. n.a. 0.0 0.0 0.2

D. Other Ethnics:

English only .......... 53.3 53.8 n.a. n.a. 55.9 50.7 44.4French only ........... 5.0 3.7 n.a. n.a. 5.3 5.6 10.8

English and French ... 34.5 35.9 n.a. n.a. 36.1 37.2 31.4Neither ............... 7.3 6.7 n.a. n.a. 2.7 6.6 13.4

* Not available.

1 Bell Telephone Company of Canada, Mont-

real Classified Telephone Directory, July, 1938,

pp. 108, 223. I am indebted to G. L. Long, his-

torian, Bell Telephone Company of Canada, for

making this volume available. Yellow Pages,

Montreal, September, 1964, pp. 291, 552.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 9: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 9/17

BILINGUALISMN MONTREAL 17

Canadian separatism reached its apex.Finally, if a segment of the populationknows English but refuses to speak it, then

for many functional purposes such people

are equivalent to those who know onlyFrench.

Consideringthat in 1961 French was themother tongue of nearly two-thirds (65

per cent) of the metropolitan populationand English the mother tongue of littlemore than one-fifth (22 per cent) of the

population, it is clear that the position ofthe French language is far from being as

strong as we might expect simply on thebasis of the composition of Montreal's

population. The number speaking Frenchonly amounts to 60 per cent of the 1.4 mil-lion persons whose mother tongue isFrench (FMT). The monolingual Englishpopulation by contrast, amounts to 93 percent of the 495,000 persons whose mothertongue is English (EMT).316 For 1961, itis estimated that 28 per cent of EMTBritish males were bilingual. On the otherhand, about 48 per cent of the FMT

French males were bilingual.'7 Equally lop-sided differencesare found for females, al-

though they are of a lower magnitude. In

other Canadian cities, where the relative

numbers with French and English mother

tongues are reversed, French fares far

worse as the lesser languagethan does Eng-lish in Montreal. In Moncton, N.B., for

example,whereFrenchis the mothertongueof about a third of the population andEnglish the mother tongue of nearly all

the remainingtwo-thirds,only about 4 percent of EMT British males have learnedFrench, whereas nearly 90 per cent of FMT

French males were bilingual in 1961.

INTERGENERATIONAL MAINTENANCE

If bilingualismis to lead to the demise ofone of the languages, then the secondlanguageof some bilingual parents must be

passed on as the mother tongue of theirchildren. If bilingualism is to be the endproduct of linguistic contact, then themother tongue of bilingual parents must be

passed on as the mother tongue of theirchildren. To simplify the problem, we willconsider three elements in our model oflinguistic maintenance.

First, the proportion in each mother-tongue group that is bilingual clearly in-fluences what we might call the "exposureto risk." We observed for the two leadingethnic groups that a far larger proportionof FMT residents learn English than EMTresidents learn French. For all fathers of

small childrenin Montreal-Verdunn 1961we estimate that about 35 per cent of thosewith EMT have learned French, whereasabout 49 per cent of FMT fathers areestimated to be bilingual.'8 In the previoustwo decades these differences in exposureto risk were even greater. French is sub-ject to a far greater "exposure" than isEnglish in the sense that proportionatelymore of the French are potential spreadersor carriers of the English language thanthe Englishare of the French language. Therelative degree of bilingualism among thedifferent speech groups is critical, since itis reasonable to assume that only bilingualparents face an option in the tongue theypass on to their offspring.

Second,we should considerthe degree to

which each bilingual group passes on the

acquired non-mother-tongue language totheir offspring, what we might call the

"rate of intergenerational anguage-switch-ing." It is most important that the actual

behavior of bilinguals with respect to lan-

guagetransferbe distinguished rom the rateof bilingual exposure. The sheer fact that

the FMT group is more bilingual than is

"6This does not mean that 93 per cent of the

EMT population speak English only. Many EMT

persons are bilingual but are replaced in this figureby people with other motjier tongues who, of the

two official languages, speak only English.

17 Lieberson, "Language Questions in Censuses,"

op. cit.

18 Estimates are based on the official-language

distribution of the "other" mother tongue popula-

tion in 1931 (most recent year for which such

data are available) and the assumption that allEMT or FMT persons in Montreal-Verdun retain

speaking ability in their mother tongue.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 10: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 10/17

18 THEAMERICANOURNALOF SOCIOLOGY

the EMT group does not necessarily mean

that a larger proportion of French bilin-

guals pass on a second language to theirchildren. In fact, it is possible for the two

sets of rates to be reversed so that there isa net decline in the language of the lin-guistic populationwith the smaller degreeofactual bilingualism. This can occur if pro-portionately more of its bilingual members

than of the bilingual component of theother language group actually pass on a

second language.Third, we should examine the fertility

ternal and international migration. How-ever, the three elements described above-exposure to risk rate of intergenerationallanguage-switchingand fertility-appear to

be the major factors to consider in theMontreal situation.

If we compare the mother-tongue dis-tribution of small children (Table 3, cols.

[1]-[3]) with that of women in the child-bearing ages (Table 3 cols. [4]-[6]), wefind that the higher degree of bilingualism

among the FMT population has not led toa net switch to English among the children.

rates separately for the monolingual and

bilingual subpopulations of each mother-tongue group. There can be a higher rate

of language-switchingamong the bilingual

segment of a mother-tonguegroup without

a decline in the language if this segment's

fertility rates are relatively low comparedwith those of the monolingualmembers.In

this sense, we must always distinguish be-

tween aggregateand individual assimilation.It is possible for individuals to assimilateand yet for their group to maintain itself

or to expand in size.'9There are additionalvariablesthat might

be added to this simplifiedanalyticalmodel,

such as differentials in mortality and in-

In all periods, a larger proportionof small

childrenthanof womenin the child-bearingages have French as their mother tongue.

In 1951, for example, 67.9 per cent of the

women had French as a mother tongue,

whereas 70.1 per cent of the childrenwere

reportedas having French. The position of

the English language is somewhat incon-

sistent, in some decades being slightly bet-

ter than its rank among women of child-

bearing ages would suggest and in other

instancespoorer (cf. Table 3, cols. [1] and

[4]). The erosion of other mother tongues

is quite apparent; in each decade their

over-all position among small children is

weakerthan the position amongadult wom-

en. The net results of our intergenera-tional analysis indicate that French is not

TABLE 3

PERCENTAGE DISTRIBUTION OF MOTHER TONGUE OF CHILDREN UNDER 5 (ACTUALAND EXPECTED) AND OF WOMEN IN THE CHILD-BEARING AGES

MONTREAL AND VERDUN, 1941-61

CHILDREN UNDER 5 YE ARS CHILDREN UNDER 5 YEARS

(ACTUAL) , (EXPECTED)

English French Other English French Other English French Othet

(1) (2) (3) (4) (5) (6) (7) (8) (9)

1941 19.8 73.0 7.2 23.9 66.1 10.1 20.4 70.9 8.71951 . 24.8 70.1 5.0 23.0 67.9 9.0 21.2 70.0 8.81961 16.5 70.7 12.9 17.3 68.4 14.3 14.5 70.8 14.71961 (MA)*. 22.6 68.3 9.0 22.0 65.9 12.1 20.2 68.5 11.3

* Entire CensusMetropolitan Area.

Note: Cols. (7)-(9) basedon ethnic originof children underfive and the relationshipbetween ethnic originand mother tonguefor all males in the area.

19 Stanley Lieberson, Ethnic Patterns in Ameri-can Cities (New York: Free Press of Glencoe,

1963), pp. 8-10.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 11: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 11/17

BILINGUALISMN MONTREAL 19

merely holding its own but is actually

gaining between generations.Before interpreting this finding, it is

necessary to consider the differentials in

fertility between language groups-par-ticularly since the French have higher

fertility rates than the non-French ethnic

component of Montreal.20As a crude ap-

proximationwe have determinedthe mother

tongues we would "expect" among chil-

dren under five years of age knowing their

ethnic origins and the cross-tabulationbe-

tween mother tongue and origin for males

in the metropolitan area. Using Wester-

gaard's method of standardization,we de-

rive the "expected" mother-tongue distri-bution of children under five given in

columns (7) through (9) of Table 3.

Several comments are called for before

discussing the results. First, we standard-

ized on the basis of males rather than fe-

males because ethnic origin in the case of

mixed offspringis traced through the male

lineage in the Canadian census enumera-

tions. Second, the cross-tabulationbetween

origin and mother tongue is not available

for specific age categories. Consequently,

the standardization rates incorporate to

some extent the very data examined for

small children.Since childrenunder five do

not comprisea large segmentof the popula-

tion, we are not too greatly hamperedby

this departurefrom the ideal.21 Finally, the

available data do not allow for the control

of differences in fertility between mono-

linguals and bilinguals within each ethnic

population. This has important implica-tions, which we will discuss shortly.

Controlling for fertility differences be-

tween ethnic groups, we find the French

language has about held its own (Table 3,

cols. [2] and [8]). In both 1951 and 1961,

the position of French among children iswithin 0.2 per cent of the figure we wouldexpect on the basis of the ethnic origins ofthe youngsters. In 1941, the number of

FMT children is even greater than whatwe would expect on the basis of ethnicfertility differentials.The gain for French

in the two most recent decades among chil-dren comparedto its position among adultwomen, however, is not due to a nettransfer to French among the bilingualparents of Montreal. Rather, it appearslargely to reflect the higher fertility ratesof ethnic groups with sizable numbers ofnative French-speakers.

By contrast, ethnic fertility differentialsare clearly unfavorableto the maintenanceof English. In each period the percentageof children we would expect to have Eng-lish as a mother tongue falls below thepercentage of women in child-bearing ageswith this mother tongue (cf. Table 3, cols.[4] and [7]). In the last two decades,how-ever, the English language has more thanheld its own when we consider the lower

fertility of ethnic groups with large num-bers of EMT adults (cf. Table 3, cols. [1]and [7]). In 1951, for example, takingfertility into account, we would expect 21.2per cent of children to have English as amother tongue whereas actually 24.8 percent were reported as learning this lan-

gauage first. The excess was more than

enough to compensatefor lower fertility in

1951 and results in the percentageof small

children with English exceeding the per-

centage of women with this language(respectively, 24.8 and 23.0 per cent). The

results for 1961 are inconsistent, dependingon whether one includes the entire metro-

politan area in the analysis. What is clear,however, is that the net intergenerationalswitching has been favorable to Englishbut that much or all of the gain is wipedout by the lower fertility rates.

Although the data relevant to the ques-

tion of whether fertility differentialsaid or

handicap other mother tongues are incon-sistent (Table 3, cols. [6] and [9]), there

20 The French have higher fertility even with

education and income held constant. See Enid

Charles, The Changing Size of the Family in Can-

ada (Ottawa: Dominion Bureau of Statistics,

1948), p. 105.

' Moreover, where there are any trends throughtime, the child-bearing ages would probably fall

close to the mean for each ethnic group.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 12: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 12/17

20 THEAMERICANOURNALOF SOCIOLOGY

is little doubt that other languages in

general fail to hold their own. Whether the

mother tongues of children (col. [3]) are

compared with those of women in the

child-bearingages (col. [6]) or with thoseexpected among children on the basis of

ethnic differentials in fertility (col. [9]),the results clearly show a drop in other

mother tongues among small children. If

anything, the extent of this drop is prob-

ably understated, since many small chil-dren were either born in a country where

English was not spoken or were born in

Canada of relatively recentimmigrantswhohad not yet sufficiently mastered one of

the officiallanguages to use it for speakingto their children.

It is very difficult to estimate fertility

differences between the linguistic sub-

populationswithin each ethnic group.Over-

all, on the basis of indirect evidence, we

wouldbe inclined to guess that the bilingual

segment has lower fertility rates. This

guess is based on the following facts:First, Enid Charles finds an inverse asso-

ciation between fertility and income with-

in each major ethnic population.22Second,

there is some evidence that bilinguals tend

to have higher incomes than monolinguals

within both the French- and the English-

speaking populations.23This would sug-

gest, although by no means conclusively,

that the bilingual segments of Montreal's

FMT and EMT populations have lower

fertility than their monolingual com-

patriots. Under such circumstances, the

exposure-to-riskrates overstate the dangerto the mother-tongue populations of bi-

lingualism.The intergenerationalanguage-switching

rate is more elusive than the two elements

of the model considered thus far. Indeed,

we do not have available the basic data

needed for determiningthe switching rates

for the bilinguals of each mother-tonguegroup. We do wish to note, however, that

if the languages are maintained between

generations in a simple two-language-

contact setting with some switching be-tween generations,then if fertility is equalit can be shown that actually the bilinguals

of the group with lower exposure to risk

are experiencing a higher rate of inter-

generational language-switching than is

the groupwith the largerpercentageof bi-

lingual parents. It is reasonableto assume

that, in a bilingualsetting such as Montreal,some parents do in fact pass the secondlanguage on as their children's mother

tongue. Under such circumstances, if theFMT population is more bilingual than

the remainder of the population and if

switching is equal in each direction, then

it follows that the rate of switching is

higher for the bilingual non-EMT popu-

lation than for the bilingual FMT popula-

tion.One important consideration in inter-

preting the maintenance of both English

and French throughthe

generationsis the

fact that men in Montreal are more bi-

lingual than women. This sex difference

means that many of the bilingual members

of each ethnic group are married to mono-

lingual mates who share only the same

mother tongue. Insofar as this occurs, the

high incidenceof bilingualism s reduced as

a contributing force to language transfer,because only one of the mates can use the

second language with their children. Sex

differencesin bilingualismwithin a largelyendogamous population help to maintain

the common mother tongue of the mates

in the next generation,since this would be

the one language that both parents could

use with the children.Of course, marriages across mother-

tongue lines are considerablymore complex

in their possibilities. Marriages between

EMT and FMT mates will often involve a

cross between Catholic and non-Catholic.

If the ties of the Roman Catholic church

are strong enough so that frequentlymates

' Charles, loc. cit.

3 In a current study of want ads in Montreal

newspapers, jobs that require bilinguals tend tobe concentrated in categories that offer higher

income.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 13: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 13/17

BILINGUALISMN MONTREAL 21

will agree to a Catholic education forprospective offspring, intermarriage willnot harm French-languagemaintenance in

Montreal, since the Catholic school system

is overwhelmingly French-speaking.24

COHORTS

Thus far our demographic examinationhas focused on intergenerational dimen-

sions of bilingualism, namely, the languagetransferred by bilingual parents to theiroffspring.At this point, we turn to viewingsecond-language-learning in connectionwith the life span of residents of Montreal.The terms "cohort," "longitudinal," and

"generational" will be used interchange-ably to describe the approach employed.As opposed to the more commonly encoun-tered cross-sectional analysis, a longi-tudinal study allows us to determine truesocial change since the linguistic behaviorof a population is actually traced throughtime.25 By considering where in the life

span second-language-learningends to oc-cur, we may infer the social condifionsleading to bilingualism.

There are several serious difficulties in-volved in tracing the bilingualism of agegroups through time. First, we have no

way of directly controlling for the foreign-born population in Montreal and cannotreadily prevent immigration between dec-ades from affecting the analysis. Similarly,

lack of available data prevents the appli-cation of controls for emigration and in-ternalmigration. The results presentedare,

therefore,crude cohorts which describe the

process of linguistic change by age groupsbut which do not easily allow for the decom-position of these changes into the contribu-tions that selective migration and mortalitymay have contributed to the results. Withrespect to mortality differentials,however,it is unlikely that these would significantlyalter the results. We shall have more tosay about the influence of immigrationlater.

Shown in Table 4 are the bilingual pro-

portions by age and sex or the four dec-ades, 1931-61. For 1941-61, the analysisis based upon the combined cities ofMontreal-Outremont-Verdun.Since data

are not available for Outremont in 1931,the 1941 figures were recomputed for theMontreal-Verdunarea alone.

Mlale Cohorts.-If we assume that theproportion of boys who are bilingual at

the age they usually begin school fallssomewherebetween the proportionsfor thepopulation under five and five throughnine, then we would estimate that lessthan 10 per cent speak both English andFrench by the time elementary school isbegun. This means that the linguisticheterogeneity of Montreal has a limitedeffect in introducing bilingualism among

boys at an early age. What appears tooperate is a steady push in the learning ofa second official language during the ages

of mass education, some additional gainin bilingualismin the early adult ages, andthen a steady decline in bilingualismthroughthe remainingyears. Tracing male

age groups between 1931 and 1941, forexample, we find 4.1 per cent of children

under five are bilingual (col. [1]), butten years later boys ten through fourteenyears of age are 22.2 per cent bilingual

(col. [2]). Similarly, 18.2 per cent of boys

five through nine are bilingual in 1931,

but more than half (51.4 per cent) arebilingual ten years later. This result holds

24Only two of the fourteen elementary-school

districts of the Montreal Catholic School District

are English. See Bureau of Statistics, School Di-

rectory, 1963-1964 (Montreal: Montreal Catholic

School Commission, n.d.).

'Among recent examples of the application of

cohort analysis to empirical data are Pascal K.

Whelpton, "Trends and Differentials in the Spac-

ing of Births," Demography, I (1964), 83-93;

Hope T. Eldridge, "A Cohort Approach to the

Analysis of Migration Differentials," Demography,

I (1964), 212-19; Beverly Duncan, George Sabagh,

and Maurice D. Van Arsdol, Jr., "Patterns of City

Growth," American Journal of Sociology, LXVII(January, 1962), 418-29; N. B. Ryder, "The In-

fluence of Declining Mortality on Swedish Repro-

ductivity," in Current Research in Human Fertility(New York: Milbank Memorial Fund, 1955), pp.

65-81.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 14: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 14/17

22 THEAMERICANOURNALOF SOCIOLOGY

up as well for the 1941-51 and 1951-61comparisons (cols. [3]-[4], [4]-[5]). Inaddition to the rapid buildup of bilingual-ism through the teens, there is some incre-

ment in the early adult ages.This surge in bilingualism tapers off for

each cohort somewhere n the twenties andearly thirties and is then followed by anactual net decline amongmen in the middleand later years of their life. Men thirty-five and older in 1931 all show declinesin bilingualism duringthe followingdecade

(cols. [1] and [2]). This interdecade de-

are unable to speak both official languages.The decline that follows the school yearsindicates that economic and occupationalforces in combination with internal and

international migration fail to fully sup-port or maintain bilingualism among a fairnumber of males. Once the supports of edu-cation and the early years in the laborforce are past, ability in both official lan-guages begins to decline. It is as if we wereto examine knowledge of algebra by agecohorts over time. We would expect to finda rapid rise through the school ages, fol-

cline starts even earlier in the 1941-51span, with men twenty-five and over in1941 showing declines in bilingualism ten

years later. Indeed, the age at which thedecline in bilingualism begins is evenearlier by 1961. As we shall see, the de-cline in the age at which male bilingualismbegins to erode is explainable in terms ofimmigration.

These results suggest that the educa-tional system of Montreal, which requires

courses in the official language not used asthe language of instruction in the school,meets with success. This success is far

from complete,since in most decades abouthalf of the boys fifteen through nineteen

lowed by a decline after the educationalperiod, since many find little or no needfor algebra in either their occupational or

their social worlds. In similar fashion,through a fair part of the adult male span,more bilingual speakers lose their knowl-edge of the second language than mono-linguals acquire a knowledge of thesecond official language.

Female Cohorts.-Further support forthis interpretation comes from examina-tion and comparisonof bilingualism amongthe female age cohorts (Table 4, cols.[6]-[10]). First, we observe that the de-

gree of bilingualism among girls under 15is very similar to that found for boys of

TABLE 4

PERCENTAGE BILINGUAL, BY AGE AND SEX, MONTREAL AREA, 1931-61

MALES FEMALES

AGE Montreal-Verdun Montreal-Outremont-Verdun Montreal-Verdun Montreal-Outremont-Verdun

1931 1941 1941 1951 1961 1931 1941 1941 1951 1961

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

0- 4... 4.1 5.7- 5.7 3 3 2.5 4.0 5.6-- 5.7\ 3.4 2.55- 9 ... 18.2 11.3-- -11.5\\ 9.7 9.9 18.0\ \11.5- 11.8 \ 9.7\ 9.6

10-14 ... 43.4\ -22.2 22.6\ \20.5 \22.4 41.4 21.9- 22 3\ 20.1\ 21 9

15-19 ... 62.4 51.4--51.7 50.6 49.6 54.7 \43.1 43 5\ 44.5 46.720-24 ... 67.2 67.1- 67.2k 64.9, 59.4 53.3N s 51.5- 51.7-' 48.2\ s

44.425--34 .. 61.9 > 68 8- 68.8 68.8 59.7 49.0 47.8- 48.1 ss 47.8 > 41.135-44 ... 62.2 - 63.6 63.7 - 68.1 65.3 44.5s 40.9 '-41.2,, 45.2 45 545-54. . 59.3 60.3 60.3 -62.7 -. 63.6 41.6 35.6- 36.0- -J37.41 J-42 655-64.. 57.4 53.7--53.8 `57.3\ >57.2 37. 1, -31.2 31 6 'Jj30 8 -34 565-69.. 56.4 949.4 4.6 49.7K452.0 34.3 ' 28 0 28 5 Z26 28 57;0+ . . 51.2 42. 9- 43.3 x 42.2 44.0 31.2 24 4- 24.7 1 23.5 24.5

___ __ _ __ _ I ____ _ _ __ _ __

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 15: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 15/17

BILINGUALISMN MONTREAL 23

the same ages. With the exception of ten-through fourteen-year-old girls in 1931,the sex differences are well under 1 per

cent. This suggests that the experiences

of boys and girls are very similar at theseages so that the net influences of homelanguage, neighbors, mass media, play-mates, and early years of formal second-language instruction in the schools arenearly identical. Beginning with the lateteens, at the age when formal educationends for many and participation in thelabor force begins, we find increasing dif-

ferences between the sexes in their bi-lingualism. By the early twenties, the sex

differences n bilingualismare considerable.In 1931, for example,43.4 and 41.4 per centof the males and females ten through four-teen are bilingual; in 1941, 67.1 and 51.5per cent, respectively,of males and femalesare bilingual. Percentage-point differencesof about 15 are also found for 1951 and1961 between males and females in thetwenty through twenty-four age category.

The influence of male participation in

the labor force and female withdrawal into

the home and child-rearing can also beseen in the points at which a net declinein bilingualism occurs as well as in thelower increases in bilingualism amongwomen. In all three interdecade compari-sons we find that declines in the bilingualpercentageoccur at earlier ages for womenthan for men. This is most striking forthe 1931-41 comparisons. Not until wereach men thirty-five through forty-four

in 1931 do we find a cohort that declinesin the succeeding ten years (62.2 in 1931

and 60.3 in 1941). By contrast, femalesfifteen through nineteen and older decline

in bilingualism by 1941.26 These differen-

tials are compatible with the contention

that the main supports of bilingualismare

school and occupational systems, althoughthe reader should be reminded that the

evidence is hardly incontrovertible.

Inmigration.-Unfortunately, limitations

in the available data preclude an analysisor weighting of the specific contributionsmade by such diverse institutional and

demographic forces as education, occupa-tional demands, internal migration, andimmigration n influencing the bilingualismof cohorts. It is possible, however, to gaugethrough indirect means the influence of im-migration in interdecade changes in bi-lingualism. There is little need to considerthe impact of recent immigration on bi-lingualism among the younger age groups,since the rapidly rising bilingualism ratesrun counter to any expectation we might

have on the basis of the low bilingualismof recent immigrants.

The decade between 1931 and 1941 af-fords the easiest test, since there was verylittle net immigration to Montreal-Verdun.For some age categories in 1941, less than1 per cent of the population were foreign-bornwho had arrivedin the preceding dec-ade. Recent immigrants (since the 1931census) were strongest among the thirty-five through forty-four-year-oldpopulation,comprising2.2 and 2.8 per cent, respective-ly, of males and femalesin 1941. Even if noimmigrants in the preceding decade hadbecome bilingual-which is far from thecase-their number would have been toosmall to account for the interdecade de-clines in bilingualism that we observe in1941 among men who were thirty-five andover in 1931 and among women who werefifteen and older. We can conclude that

the inferences based on cohort changes be-tween 1931 and 1941 are substantially un-altered by immigration during the decade.

Heavier immigration since World WarII definitely influences the cohort patternsdescribed earlier. Among both men andwomen in the age category thirty-fivethrough forty-four in 1961, about 20 per

cent were immigrantswho had arrived in

Canada during the preceding ten years.

For both 1951 and 1961 we have estimated

the age-specific bilingual percentages, ex-cluding for each year immigrants first ar-

26

Here and elsewhere, graphs have been em-ployed to compare cohorts over time that are in

different-size age categories in the two periods.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 16: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 16/17

24 THEAMERICANOURNALOF SOCIOLOGY

riving in the preceding decade ("recentimmigrants").27 We shall examine theserates in order to determine whether theinferences made earlier are valid after the

effect of immigration is eliminated.The results, shown in Table 5, indicate

that no substantial changes need be madein the cohort analysis presented earlier. In1951 we find that the decline in bilingual-ism begins among men who were forty-fiveand older in 1941 (cols. [1] and [2]).Among women, the decline now begins

steady decline is observed among womenthirty-five and over in 1951.

In brief, the existence of unlearning isstill found among the populationwho were

not immigrants to Canada in the interven-ing decades. The process of immigrationtends to lower the age at which bilingualismstarts to decline. In other words, the effectof large-scale immigration on cohorts isthe same as if bilingualism were to start itsdecline at earlier ages. The decline amongmen is more of a middle-age phenomenon,

among those twenty and older in 1941(cols. [5] and [6]). In 1961, the declinefor men starts with those who were thirty-five and over ten years earlier (cols. [3]and [4]). The results for women are some-

what inconsistent; there is a decline forwomen twenty throughtwenty-fourin 1951but a rise between 1951 and 1961 amongwomen twenty-five through thirty-four atthe beginning of the period. However, a

while women tend to begin their decline inbilingualism at earlier ages. This result is,of course, compatible with the earlier in-terpretation of the institutional forces thatmaintain bilingualism in Montreal.

COMMENT

Racial and ethnic contact is frequentlyaccompanied by the confrontation of peo-ples who speak different tongues. In theUnited States, with a few exceptions, wehave been able to take more or less forgranted the linguistic outcome of contactswith non-English-speaking peoples. Al-though the experienceof the United Statesis not unique, there are far more complexlinguistic situations in many parts of theworld. Indeed, Canada is relatively simple

TABLE 5

PERCENTAGE BILINGUAL FOR SELECTED AGE GROUPS, BY SEX, EXCLUDING RECENT

IMMIGRANTS, MONTREAL-OUTREMONT-VERDUN, 1941-61

MALES FEMALES

AGE

1941 1951* 1951 1961t 1941 1951* 1951 1961t

(1) (2) (3) (4) (5) (6) (7) (8)

10-14 .... 22.6 20.5, 22.3 20.115-19 .... 51.7,X 50.6 s O43.5 44.520-24 67.2 66.8 64.9 64.7 51.7 '49.5 48.2 49.125-34 .... 68. 8 72.2 68. 8 67.6 48. 1 50 0 47.88 46.035-44 .... 63. 7 71.2 68.1 - 71.6 41.2 - 46.4 45.2 . 49.045-54 .... 60. 3 64. 3 62.7 - 67.5 36. 0 38. 1 37.4 44.7

55-64 .... 53. 8 -58.0 57. 3 59.0 31.6 31.1 30.8 35. 765-69 49.6 50.1 49.7 53.0 28.5 26.7 2 6. 5 29.370+ 1 C42. 5 44. 5 23.7 24.9

* Excluding immigrants since 1941.

t Excluding immigrants since 1951.

I The bilingual percentage for the population,excluding recent immigrants, was obtained through

indirect standardization. Using the component of

recent immigrants in each group as weights, the

ratio of actual and standardized bilingual percent-

age was applied to each specific age category to

estimate the bilingual percentage of the popula-tion who had lived in Canada in the preceding

decade.

This content downloaded from 152.74.16.35 on Sun, 23 Jun 2013 16:23:56 PMAll use subject to JSTOR Terms and Conditions

Page 17: AJS Bilingualism in Montreal a Demographic Analysis

7/28/2019 AJS Bilingualism in Montreal a Demographic Analysis

http://slidepdf.com/reader/full/ajs-bilingualism-in-montreal-a-demographic-analysis 17/17

BILINGUALISMN MONTREAL 25

compared to such nations as Nigeria, India,or South Africa.

The examination of linguistic trends inMontreal indicates a process quite con-

trary to the view of the city as a greatmixer and melting pot of diverse cultures.

Unlike most American cities, where popu-lations with diverse linguistic origins havemoved toward a monolingual status in amatter of a few generations, Montreal hasmaintained an equilibrium. This equilib-rium is similar to that of Switzerland inthe sense that it is based on dynamicdemographic forces that tend to counter-balance each other.28 It is a precarious

equilibrium, since the exposure to risk isgreater among the French ethnic popula-tion than among the British.

If we make the assumption that mostpeople will not really master a secondtongue unless it is learned at a relativelyearly age, then second-language-learningsless of a threat to the mother tongue thanmight otherwise be the case, since Englishand French languages are able to holdtheir own as the first language of children,and since muchof the bilingualism n Cana-da does not occur in the very early ages oras a result of informalsocial contacts. Key-fitz has pointed to the handicaps that bi-lingual French Canadians face in the upper

echelons of management or in white-collaroffice settings because of their lack offluency or ease in English, which is oftenthe languageused.29

The constant replacement of a popula-tion through the introduction of new gen-

erations means that social change mayoccur without any individual changes.30However, we should not lose sight of thefact that generations may be instrumentsof conservation. Insofar as new members

are socialized by older members, then it isequally possible for the system to main-

tain itself over time indefinitely as thenewcomerswho gradually replace the older

members are socialized.8' In the case ofMontreal, we witness a social process of

second-language-learningin each genera-

tion that does not lead to change throughthe years in the linguistic structure of

Montreal. From the perspective outlined

earlier, transfer from parents to childrenand the stage in the life span at which

bilingualism occurs are both critical for

understanding the outcome of languagecontact.

UNIVERSITYOF WISCONSIN

28 For an excellent analysis of the demographicfactors, see Kurt Mayer, "Cultural Pluralism andLinguistic Equilibrium in Switzerland," reprintedin Joseph J. Spengler and Otis Dudley Duncan(eds.), Demographic Analysis (Glencoe, Ill.: FreePress, 1956), pp. 478-83.

'2Nathan Keyfitz, "Canadians and Canadiens,"Queen's Quarterly, LXX (Summer, 1963), 163-82.

30Karl Mannheim, Essays on the Sociology of

Knowledge (New York: Oxford University Press,1953), pp. 276-320.

" See George Simmel, "Social Interaction as theDefinition of the Group in Time and Space," in

Robert E. Park and Ernest W. Burgess, Intro-duction to the Science of Sociology (Chicago:

University of Chicago Press, 1921), pp. 348-56.