Chapter 4, Choosing a Code

Diglossia

Psichari (1928) - In what is perhaps the earliest use of diglossia, this writer refers to the situation of Greek at the end of the 19th century, describing Greece as "a country that doesn’t want its language." This observation refers to the fact that while Dimotiki ("popular language, " "people’s language") is used as the everyday medium of communication, Katharévusa ("pure language") is used for writing, and reflects Classical Greek more than the popular form.

Marçais (1930) - This writer described the situation in the Arabic world in the thirties, when the gulf between spoken Arabic dialects and the classical standard was particularly large. Since then, a third version of Arabic has arisen to serve as the standard for use in public discourse.

Ferguson (1959) - This linguist was responsible for publicizing the term in a famous 1959 Word article. Swiss German, Haitian Creole

Fishman (1967) - The widespread nature of Paraguayan bilingualism caused Fishman to hypothesize that diglossia could occur in any situation where two language varieties, even unrelated ones, are used in functionally distinct ways.

Ferguson's definition - the side-by-side existence of two structurally and historically related language varieties (a High variety and a Low variety, referred to as H and L) throughout a community, each of which has a distinct role to play (examples found in Greece, Egypt, Haiti, and Switzerland)

Defining Criteria of Diglossia

Function - H is the more elegant, formal variety. L is used for less politically important functions.

Prestige - Attitudes toward H are more positive than towards L. H is the prestigious variety and L is the stigmatized variety.

Literary Heritage - H is associated with a long literary tradition. H is always used in writing. L fulfills few written functions. It may be found in cartoons or in the speech of characters in novels.

Acquisition - L is always acquired as a first language. H is always learned in a formal, educational setting.

Standardization - Dictionaries and grammars document the form of H. L usually has no such support.

Stability - Diglossia is a long-lived phenomenon. Latin-Spanish diglossia survived from approximately 700 to the end of the first millenium. H and L borrow from one another, although L forms are shunned when using H.

Grammar - The morphology of L is often simpler than that of H. Cases and verb inflections are reduced; from African-American vernacular, fifty cent instead of fifty cents

Lexicon - A striking feature of diglossia is the existence of paired lexical items, where L and H have different terms for the same object; from Paraguayan Guaraní, silla instead of apyka (chair)

Phonology - H preserves the underlying phonological system, and L diverges from it, typically having evolved away from the classical form over many hundreds of years; from Vulgar Latin, specla instead of specula (mirror)

Fishman’s reformulation

 

+Diglossia

-Diglossia

 

+Bilingualism

Everyone in a community knows both H and L, which are functionally differentiated. (Haiti)

An unstable, transitional situation in which everyone in a community knows both H and L, but are shifting to H. (German-speaking Belgium)

 

-Bilingualism

Speakers of H rule over speakers of L (colonial Paraguay

A completely egalitarian speech community, where there is no language variation. (Humanity before the Tower of Babel)

 

Language evolution

Hudson (1990) has pointed out that Fishman’s reformulation of the concept of diglossia is problematic, because the direction of language evolution in a classic diglossic situation is opposite to that in the case of widespread bilingualism

Ferguson’s diglossia: L/H Æ L

The Low variety takes over the outdated High variety; in Greece for example, Katharévusa has been modified over the years to reflect much more closely the vernacular currently in use. The same phenomenon has occurred in the Arabic world.

Fishman’s diglossia: L/H Æ H

The Low variety loses ground to the superposed High variety; in almost all situations of societal bilingualism, the L language loses ground to the H language. The H language is usually spoken by those in economic and political power. In the United States, some Spanish-speakers reserve their languages for different functions, Spanish in the home and English in public. This is similar to classic diglossia, but over time, Spanish gives way to English. Children end up learning the H variety and leaving the L variety behind. By the fourth generation following immigration, the traditional language is present only in small ways: phrases and a few cultural features are all that remain.

 

Bilingualism

Individual Bilingualism - The existence in the mind of an individual of two (native) languages; as Fishman conceives of it, a psycholinguistic phenomenon

Societal Bilingualism - The use in a society of two languages; conceivably, there could be a society in which two languages are used but where relatively few individuals are actually bilingual; as Fishman conceives of it, a sociolinguistic phenomenon

Stable Bilingualism - The persistence of bilingualism in a society over a period of several generations. Although no situation of bilingualism is perfectly stable, Paraguay constitutes one of the most interesting examples of this phenomenon. Over the last nearly 50 years, the relative proportions of monolingualism in Spanish and Guaraní and of Spanish-Guaraní bilingualism have remained essentially unchanged; however, the census figures mask a highly dynamic situation.

 

Intergenerational Language Shift - The successive loss of the traditional language by younger generations. Typical Pattern of Intergenerational Language Shift in Immigrant Communities:

first generation - Immigrants dominant in home language and know host language of host country to varying degrees

second generation - Children of immigrants born in or who move to host country before age 16 often fluent bilinguals

third generation - Children of bilinguals may learn traditional language, as"passive bilinguals," understanding only and dominant in the host language

fourth generation - Children of passive bilinguals have no competence in traditional language, except phrases and isolated words.

language choice

code-switching - changing from one language to another:

Sometimes I start a sentence in English, y luego termino en espanol.

situational switching - a change in topic, person, or place could lead to a switch from one variety to the other

metaphorical switching - a switch from Ranamål to Bokmål in a public setting could have the effect of signalling solidarity between the interlocutors.

code-mixing - speaking in one language, but using pieces from another

Shopper - ¿Dónde está el thin-sliced bread?

Clerk - Está en aisle three, sobre el second shelf, en el wrapper rojo.

style-shifting - variation within a language (changing between Standard English and African-American Vernacular)

language borrowing - permanent incorporation of words from one language into the lexicon of another language

Many English words come from other languages:

canoe

perestroika

angst

chocolate

tomato

soprano

dungaree

guitar

babble

lassez-faire

column

hound

dummkopf

algebra

teepee

lariat

voodoo

wampum

shampoo

pyjamas

beef

pork

veal

venison

 

domain analysis - Domains of verbal interaction may be defined in part by person, place, social context (situation or level of formality) and topic. Sociologists of language analyze language choice by domains. Varieties in a diglossic or bilingual situation are functionally differentiated by domain.

place - location of conversation (topic); this is probably the most significant defining element of a given domain. Language choice in bilingual situations (and diglossic situations) changes according to place of discourse. Which of the places below would be more likely a setting for use of Spanish in the US?

home (family)

neighborhood (friendship)

market (transactions)

church/temple/mosque (religion)

school (education)

factory (employment)

court (government)

person - roles of interlocutors in a conversation; these roles tend to be played in certain settings, which in turn are associated with a given language; in a situation of Spanish-English bilingualism in the US, where would the following roles be played, and in what languages would communication occur?

Person 1

Person 2

son

father

neighbor

neighbor

parishoner

priest

constituent

politician

student

teacher

client

salesperson

nurse

doctor

defendant

lawyer

coach

player

 

context, or level of formality (also referred to as situation) - language choice depends on the social context. Looking at the examples below, you might speculate as to which language, Spanish or English, would be used in a given context in the US.

evening meal (formal, at home)

lunchtime chat (informal, in government cafeteria)

pillow talk (intimate, at home)

introduction (formal, in neighborhood)

arraignment (formal, in court)

topic, or subject of conversation - topics tend to be discussed in given circumstances, and in bilingual situations, a given language will be used to discuss a given topic.

household chores

increasing vandalism

cost of vegetables

wages of sin

tomorrow's assignment

reason for tardiness

whether evidence proves guilt

documentation of deductions

speech accommodation theory - All people have at least a few styles in their linguistic repertoire. Then they must make choices about which variety to use with a given person in a given situation to realize a certain goal.

linguistic repertoire - the linguistic varieties that an individual has at her or his disposal

speech accommodation - adjustments that one makes to speech in response to the speech of another

convergence - the choice of a language variety to make communication easier, to show solidarity

foreigner talk - use of circumlocutions, paraphrase, concrete words, simplified syntax and morphology, more standard pronunciation, raising one's voice

"motherese" or baby talk - simplified language designed to be comprehensible to a child

divergence - the choice of a language variety meant to make communication more difficult, to show sociocultural distance; a native speaker may speed up, use abstract words or words that are known to be difficult to understand to nonnatives, or lower the volume of speech

An example of societal bilingualism (or Fishman-like diglossia) Frisian (summarized in Fasold 1984)

Friesland, a province in northeastern Holland, is bilingual: Dutch-Frisian. Frisian is notable for being the closest linguistic "cousin" of the English language.

Population - 550,000 (4% of the Netherlands)

Religion - First church service in Frisian held in 1915; first Frisian Bible published in 1943

Languages

Dutch 100%

Frisian 83% speak; 97% understand; 69% read; 11% write

Dutch and Frisian not mutually intelligible

Nederlands

Frysk

English

stroop

sjerp

syrup

moeder

mem

mother

vader

heit

father

maandag

moandei

Monday

Diglossia, a la Fishman

Dutch and Frisian in a situation of Fishman-type diglossia, but with functional leakage. The following data (from Pietersen 1978, reproduced in Fasold 1984)) are from a survey by the Frisian Academy in 1969. At that time, as shown below, 28% of the leaders surveyed used Frisian at home with the family. Nearly all farmers used Frisian at home. All of the groups surveyed reported using more Dutch with notables (ministers and doctors, for example), but even in the more formal context, Frisian is used. The situation is certainly not perfectly diglossic, even in Fishman’s sense, but the pattern is clearly one that relegates Frisian to more informal situations and Dutch to more formal situations. A hallmark of diglossia, according to Hudson’s interpretation of Ferguson’s conceptualization of the phenomenon, is that variation in language use depends not on who you are but instead on the social situation in which you find yourself. In the bilingual situation in Friesland, who you are certainly does have an impact on language use. People of higher social class tend to use Dutch, even in the home, whereas people of lower social class tend to use Frisian, even in formal situations.

 

Family

Notables

Occupation

Dutch

Frisian

Dutch

Frisian

Leaders

64

28

81

19

Middle

27

51

74

21

Shops

7

68

52

39

Workers

7

77

52

44

Farm labor

0

84

36

56

Farmer

1

96

33

66

 

functional leakage - partial overlap of language uses in a diglossic or bilingual situation: in Friesland, leaders use Dutch in informal situations and farmers use Frisian in formal situations.

History of Frisian Language Policy

1907, study of Frisian allowed outside of school

1937, Frisian allowed as school subject

1938, Frisian Academy established

1955, Frisian allowed in the courts (spoken only)

1972, Frisian obligatory in schools, beginning in 1980

Language Planning

Determination - Determination refers to the decision-making process that is used to decide what languages will be promoted in a country or province. Dutch is used for most public purposes, and has been historically.

Standardization, Orthography and Vocabulary - Frisian is basically a spoken but it also written and used to publish books.

Government - Government is in Dutch, Frisian allowed in courts

Education - Frisian required in primary schools; Dutch predominates

Bilingual Education

Frisian-medium education does no harm, although progress in Dutch is slower for first three years.

Frisian is a well developed small-group standard language.

Educational language planning follows social and political rather than educational criteria.

Evaluation

Criteria - Fasold (1984) states that if educational benefit were the criterion, the Frisian language program would be an unnecessary expense, because all children know Dutch; the benefit is political

Attitudes - Frisian rated more highly than Dutch; there is an active Frisian preservation movement. Nevertheless, it is important to recognize that the language is associated with rural values and the farming life. In Friesland cities such as Leeuwarden, Dutch is heavily predominant.

Comparison with US

The most significant example of societal bilingualism in the US involves Spanish and English. Think of attitudes in the US towards Spanish. Are they the same as in Friesland? Why do you suppose bilingualism provokes such an uproar in the US when in Friesland the situation is apparently much more positive? (Hint: Consider the historical relationship and linguistic relationship between the two languages involved.)

Some Data on US Spanish-English Bilingualism

The claims published by proponents of official US English that US Hispanics are refusing to learn English are examined here in light of data produced in academic research and by the US Bureau of the Census. The purpose of this analysis is to use little-understood sociolinguistic aspects of national US English-Spanish bilingualism to refute incontrovertibly the claims of many official English boosters, but at the same time to reveal the dynamic effect of change in language use on ethnolinguistic identity.

In 1990, of the 230,445,777 persons in the U.S. who were age five or

Mother tongue of U.S. citizens 5 years and over who speak a language other than English

Number of U.S. citizens 5 years and over who speak a language other than English

Number of U.S. citizens 5 years and over who do not speak English ‘very well’

Percent who do not speak English ‘very well’

Spanish

17,345,064

8,309,995

47.9%

Asian/Pacific

4,471,621

2,420,355

54.1%

TOTAL

31,844,979

13,982,502

43.9%

Table 1. US Language Use, 1990 (Persons 5 years and over: 230,445,777)

over, 31,844,979 spoke a language other than English at home (see Table 1). Of these, 13,982,502, or approximately 6% of the U.S. population reported not speaking the English at the level ‘very well.’ The census bureau reports that over 75% of nonnative English speakers claim to speak English at least ‘well’. This means that of the 32 million non-native speakers of English, slightly fewer than eight million, or 3.5% reported speaking English less than ‘well.’ Even someone who reports that his or her English is only ‘fair’ can hardly considered to be a non-English-speaker, so this method of determining acceptable English proficiency is conservative. Nonetheless, even using this conservative estimate, 96.5% of the country speaks English ‘well’ or ‘very well.’

Within this national context, the figures in Table 1 show that while the proportion of US Hispanics who report speaking English ‘very well’ is somewhat higher than that of the total population of non-English-mother-tongue Americans, a higher proportion of the US population of Asian/Pacific origin is of limited English proficiency (as defined by this overly conservative method). The figures in Table 1 also reveal that although speakers of many other languages were also living in and immigrating to the United States, speakers of Spanish constituted the overwhelming majority of individuals claiming a language other than English as their mother tongue.

Bills, Hernández-Chávez and Hudson have refined a number of relevant measures which simplify the job of understanding language shift. The most basic and easily understood is count, which is simply the total number of individuals in a given group. In Table 2 the U.S. and U.S. Hispanic counts are presented. The figures are indeed striking. The historic increase in the numbers of United States Hispanics that occurred during the last decade was actually eclipsed in some respects by the increase in Hispanics during the seventies, which alerted the supporters of official English to the challenge that their language faced.

In 1970, the total U.S. population count was 203,302,031, and Hispanic density, defined by Bills, Hernández-Chávez and Hudson as the proportion of the population that is of Hispanic origin, stood at only 3.9% (see Table 2). Just over 12% of those Hispanics had immigrated to theUnited States during the previous decade. By 1980, of the total U.S. population of 226,545,580, 6.4% was Hispanic. The Hispanic population had increased by 5,536,017 to 14,608,673.

Census Year

United States Count (USC)

Hispanic Count

(HC)

Hispanic Density (HC/USC)

Hispanic Immigrant Count (HIC)

Hispanic Immigrant Density (HIC/USC)

1970

203,302,031

9,072,602

3.9%

1,104,500

0.5%

1980

226,545,580

14,608,673

6.4%

1,408,300

0.6%

1990

248,709,873

21,900,089

8.8%

2,799,400

1.1%

Table 2. U.S. Hispanic Count and Density, 1970-1990

The data on the increase in the U.S. Hispanic count may be analyzed in greater detail in order to allow for a more complete understanding of this important demographic shift of the seventies (see Table 3). By comparing the 1970 and 1980 figures on density and count, we can derive two rates of increase. The first is an increase in Hispanic count, calculated by expressing the difference between the 1980 and 1970 figures as a proportion of the 1970 count: 14,608,673-9,072,602)/9,072,602 = .61. Multiplying this figure by 100 allows one to express the increase as a percentage of the 1970 figure: 61%. The rate of increase in Hispanic count dropped to 50% in the next decade. A second rate of increase is in what Bills, Hernández-Chávez, Hudson, refer to as 'density,' that is, the percentage of the entire population that is Hispanic (see density figures in Table 2). The rate of increase in Hispanic density from 1970 to 1980 was 64%. During the next decade the rate of increase was much less at 38%.

Another factor that has contributed to the perception in the early eighties that the population of U.S. Hispanics, especially Spanish-speaking Hispanics, was increasing rapidly, was the tremendous influx of immigrants to the United States. In 1970, only 0.5% of the U.S. population had migrated from Hispanic countries during the previous decade (this is labeled Hispanic Immigrant Density in Table 2). In 1980, 0.6% of the U.S. population had migrated from Hispanic countries. Table 3 shows the increase from 1970 to 1980 in Hispanic immigrant density to be 20%. Certainly this increase was even more noticeable in border states.

Census Year

Hispanic Count

Hispanic Density

Hispanic Immigrant Count

Hispanic Immigrant Density

1970-1980

61%

64%

28%

20%

1980-1990

50%

38%

99%

83%

Table 3. Rates of Increase in U.S. Hispanic and Hispanic Immigrant Count and Density, 1970-1990 .

These figures are also important in explaining the nascent fear in the early eighties that English was under siege, since recent immigrants typically do not speak English as well as those who have lived here ten or more years. During the eighties, the increase in Hispanic immigrant count and density was even more dramatic, and lends further support to the idea that the increased linguistic evidence of Hispanic presence fueled the anti-immigrant and English-only movements of the eighties. What is especially remarkable about the data in Table 3 is the large difference between Hispanic and Hispanic immigrant rates of increase. Whereas the rate of increase in total Hispanic count and density dropped, the rate of increase in Hispanic immigrant count and density rose. To the casual observer, the effect was a notable increase in the use of Spanish in the United States during the seventies and especially during the eighties.

The above analysis of the effect of rising Hispanic and Hispanic immigrant count and density shows the basis of some of the fears of those associated with U.S. ENGLISH, but an important question has been left unanswered. Are U.S. Hispanics clinging to their mother tongue? Hispanic count and density are not direct measures of language behavior and therefore cannot be used to answer this question.

Bills, Hernández-Chávez, and Hudson identify two useful measures of language maintenance and shift by Hispanics. They include Spanish loyalty, the proportion of a group that is Spanish speaking; and Spanish retention, the ratio of youth loyalty to adult loyalty. Data on loyalty and retention based on U. S. census data are presented in Table 4. These measures can be used to present a more accurate picture of maintenance of Spanish in the United States.

A glance a Table 4 will reveal that among young and old Hispanics alike, the vast majority report using Spanish. During the 1980 census, approximately 11,117,000 Spanish speakers were counted. This figure was later revised upward to 11,549,000. Of these individuals, a total of 2,952,000 aged 5-17 spoke Spanish. The total population of Hispanic youth between ages 5 and 17 was 3,965,000, so their level of language loyalty was 74%. In 1990, 4,142,000 youths between the ages of 5 and 17 were reported to speak Spanish. Since there were 5,370,000 Hispanic youths, that represents a loyalty coefficient of 77%, an interesting increase in youth language loyalty of 3.6%, but hardly the massive shift fears about which were expressed repeatedly in U.S. ENGLISH Update.

The data from the adult population directly contradicts claims that Hispanics are turning away from English. In 1980, out of a total of 8,981,000 U.S. Hispanic adults (18 and older), 8,164,000 spoke Spanish, a language loyalty rate of 91%. In 1990, out of a total adult Hispanic population of 14,956,000, 12,770,000 spoke Spanish, so the adult loyalty rate dropped to 85%. The figures in Table 4 show that the rate of retention (referred to on the chart as ‘youth/adult loyalty’) of Spanish has actually increased by just over 10%. Since retention is the ratio of youth loyalty to adult loyalty, the increase to a large extent is due to the decrease in adult loyalty, which makes retention by the younger generation appear all the more striking. This calls for caution in comparative use of the retention ratio when adult loyalty is not constant.

Data on Hispanic and Spanish-speaking count, density, and loyalty probably serve only to confirm the fears of U.S. ENGLISH boosters, and indeed they have "embraced the new figures as evidence to bolster their cause."16 The statistics welcomed by U.S. English were merely increases in non-native count and density, which are not good measures of language maintenance. Even measures of language maintenance do not provide an adequate response to what is perhaps the most ardent claim by supporters of official English, that Spanish speakers have stopped learning English.

In order to answer the question of U.S. limited English proficiency (LEP), 1980 data are analyzed first. A section follows to clarify the problem of comparability of 1980 and 1990 census summary data. Finally,

Census Year

1980

1990

Total Hispanic Count 5 years old and over

12,946,000

20,326,000

Total Spanish Speaker Count 5 years old and over

11,117,000

16,912,000

Total Language Loyalty

.86

.83

Hispanic Count 5-17 years

3,965,000

5,370,000

Spanish Speaker Count 5-17 years

2,952,000

4,142,000

Youth Language Loyalty

.74

.77

Hispanic Count 18 years old and over

8,981,000

14,956,000

Spanish Speaker Count 18 years old and over

8,164,000

12,770,000

Adult Language Loyalty

.91

.85

Youth/Adult Loyalty Ratio

.81

.90

Table 4 - Changes in U.S. Hispanic and Spanish Speaker Count, Loyalty, and Retention, 1980-1990

1990 data are analyzed and compared with those of 1980. Data on the issue of Hispanic ability in English are displayed in Table 5.

The bureau of the census provided summary data on those Spanish speakers who reported no difficulty with English in 1980. Of the 14,609,000 Hispanics, approximately 11,117,000 age five and older reported speaking Spanish, and 2,708,000 (24% of Spanish speakers, 18% of all Hispanics, and 1% of the U.S. population) reported difficulty with English. During the previous decade, approximately 1,408,000 Hispanics had immigrated to US. Assuming that recently immigrated Hispanics have difficulty with English, by subtracting the number of recent immigrants from the total number of LEP Hispanics, a core of 1,300,000 long-term LEP (LTLEP) speakers of Spanish can be identified. To the extent that the assumption concerning the English ability of immigrants is wrong, the number of enduring monolingual Spanish speakers could be even greater. The procedure establishes a minimum limit to the count of LTLEP, the occurrence of which may be due to linguistic isolation, economic marginalization, lack of motivation, or lack of educational opportunity.

Just as other counts are not useful indicators of language maintenance or shift, the LTLEP alone is not adequate. Three indices of LTLEP density need to be derived. The number of non-immigrant Spanish-speaking Hispanics is derived simply by subtracting the number of immigrants from the Spanish-speaking Hispanic total. Dividing the LTLEP count by this figure, we obtain an index of LTLEP density among non-immigrant Spanish-speaking Hispanics of 13%. This is an important figure, for it responds to the fear that supporters of official English had in the early eighties that those who had lived for an extended period of time in the U.S. and persisted in using Spanish were rejecting English. That fear is simply unfounded. 87% of long-term U.S. Hispanic residents have no problem whatsoever with English. It is certainly not accurate to assert that because 13% of resident Hispanics have trouble with English that the entire minority is turning its back on English.

Critics of the US Hispanic presence almost unfailingly refers to all Hispanics without distinguishing on the basis of ability in Spanish, so it is appropriate that an index of LTLEP density among all Hispanics should be calculated. As can be seen in Table 5, the result is .09 or 9%. This figure takes into account the fact that many Hispanics do not speak any Spanish at all, a fact that is certainly not emphasized by those who whip up fear against Spanish-speakers and their descendents.

Finally, since critics of bilingualism publicize the putative threat that the Hispanic refusal to learn English represents to national unity, it is important to calculate the proportion of U.S. citizens who are Spanish-speaking LTLEP. The 1.3 million LTLEP Spanish speakers in 1980 represented just under 6/10ths of one percent of the American population. This, plus the newly arrived immigrants, in concrete, demographic terms, was the size of the threat that was faced in the US in the early nineties.

Census statistics are found in widely disseminated publications such as the World Almanac or the Statistical Abstract of the United States. Since the 1980 summary described Hispanic ability in English in terms of "reporting no difficulty with English," and the 1990 summary described Hispanic ability in English in terms of "not speaking English ‘very well,’" the impression caused is that Hispanic ability in English has declined over the last decade.

Census Year

1980

1990

Rate of Increase

Total U.S. Count

226,546,000

248,710,000

10%

Total Hispanic Count

14,609,000

21,900,000

50%

Total Spanish-Speaking Hispanic Count

11,117,000

16,912,000

52%

Non-Immigrant Spanish-Speaking Hispanic Count

9,709,000

14,113,000

70%

Total Immigrant Count

1,408,000

2,799,000

99%

Total Hispanic LEP Count

2,708,000

4,228,000

56%

LEP Density among All Hispanic Spanish-Speakers

.24

.25

5%

LEP Density among All Hispanics

.18

.19

6%

Spanish-Speaking LEP Density in U.S. Population

.01

.02

100%

Total Spanish-Speaking LTLEP Count

1,300,000

1,537,000

18%

LTLEP Density among Non-Immigrant Spanish-Speaking Hispanics

.13

.11

-15%

LTLEP Density among All Hispanics

.09

.07

-22%

Spanish-Speaking LTLEP Density among All U.S. Population

.0057

.0062

9%

Table 5. Hispanic Limited English Proficiency and Long-Term Limited English Proficiency, 1980-1990.

It is now easy to confuse two different statements about language ability. U.S. citizens in 1990 were asked to locate their language ability along a dimension ranging from ‘very poor’ to ‘very well.’ The 38% of Hispanics who did not choose the category ‘very well’ did not necessarily rate themselves as ‘very poor,’ ‘poor’ or even ‘fair’ (refer again to Table 1). In fact, as noted above, according to Barringer the bureau of the census reports that when the category ‘well’ is added, the number of English speakers among non-native Americans jumps to 75%. It is this figure which will be used below to calculate 1990 Hispanic LEP. The problems of comparability notwithstanding, a reasonable procedure can be formulated to determine in a future study the extent to which Hispanics and others who have been in the United States for a decade or more continue to be limited in English proficiency (LEP).

Using the census estimate that 75% of nonnative speakers of English speak the language ‘well’ or ‘very well,’ we can assume conservatively that 4,228,000 of the 16,912,000 Spanish-speaking Hispanics were LEP in 1990. Note that this figure is only roughly comparable with the 1980 census summary statistics, which reported ability in terms of having no difficulty. Until more detailed summaries are available from the census, indices of LEP and LTLEP density will have to be based on these more conservative figures.

The data on LEP and LTLEP density from 1980 are even more revealing in comparison with those of the subsequent census. In 1990, as shown in Table 5, 21,900,000 of the total U.S. population of 248,710,000 was Hispanic. The 4,228,000 Hispanics who in 1990 reported speaking English less than ‘very well’ or ‘well,’ represented only a slight increase in LEP among Hispanics (6%); however, in the United States the increase in LEP Hispanics jumped 100%.

The huge increase in Spanish LEP as a percentage of the U.S. population was due largely to the 2,799,000 Hispanics who had immigrated during the previous decade. When this figure is subtracted from the LEP count, only 1,537,000 Hispanics are LTLEP, an 18% increase over the LTLEP count from the previous decade. Reiterating the limited usefulness of count for determining language maintenance and shift, we turn to the figures on LTLEP density. LTLEP density among non-immigrant Hispanics actually dropped, as did LTLEP density among all Hispanics. Whereas LEP increased 100% in the U.S. as a whole, LTLEP increased only 9%.

Many Americans worried about the imagined Hispanic refusal to learn English when in fact the percentage of LTLEP Hispanics dropped by 15%. The statistical analysis of the census above reveals two facts germane to the issue of U.S. ENGLISH perceptions of sociolinguistic reality. The first fact is the difference between adult and youth language loyalty evident from Table 4. This interesting attitudinal change was reported in the New York Times to be documented in a study of 5000 eighth and ninth grade children of immigrants, by Johns Hopkins sociologist Alejandro Portes, who discovered high ratings of self-proficiency in English among Mexican-Americans and Cuban-Americans (85% and 99%, respectively). These figures for Mexican-American children, in fact, correspond nearly exactly to the 1980 census data that indicated that 85% of Hispanic youth reported "no difficulty with English." These figures indicate that there has been no shift from English. Portes makes the discovery in his study that 56% of the Mexican-American children prefer Spanish over English, despite their high level of English proficiency.

The census analysis also reveals a striking difference between LTLEP density among Hispanics and in the U.S. as a whole. The impressive progress in English by Hispanics resident in the U.S. for ten years or longer has been completely overshadowed by the historic increase in Hispanic immigrants. Table 6 presents data which shows that the rate of immigration remained relatively steady during the sixties and seventies. During the last decade the rate of immigration doubled. A series of events conspired to drive Hispanics to the U.S. in search of economic and political refuge. Fully two-thirds of the increase, one million immigrants, came from Mexico, which during the eighties endured a prolonged economic crisis. Political upheavals in Guatemala, El Salvador, Nicaragua, and Columbia added another 330,000. The real culprits responsible for the historic increase in the number of Spanish speakers are poverty and war.

 

Census Year

Country

1961-1970

1971-1980

1981-1990

Spain

30,500

30,000

15,800

Mexico

443,300

637,200

1,653,300

Cuba

256,800

276,800

159,200

Dominican Republic

94,100

148,000

251,800

Costa Rica

17,400

12,100

15,500

El Salvador

15,000

34,400

214,600

Guatemala

15,400

25,600

87,900

Honduras

15,500

17,200

49,500

Nicaragua

10,100

13,000

44,100

Panamá

18,400

22,700

29,000

Argentina

42,100

25,100

25,700

Chile

11,500

17,600

23,400

Colombia

70,300

77,600

124,400

Ecuador

37,000

50,200

56,000

Perú

18,600

22,700

41,300

Venezuela

8,500

7,100

17,100

TOTALS

1,104,500

1,408,300

2,799,400

Table 6. Hispanic Immigration to the United States, 1960-1990 (not included: Bolivia, Paraguay, Uruguay)