Content uploaded by Maria Polinsky
Author content
All content in this area was uploaded by Maria Polinsky on Mar 23, 2018
Content may be subject to copyright.
Heritage Languages: In the 'Wild' and in the Classroom
(Article begins on next page)
The Harvard community has made this article openly available.
Please share how this access benefits you. Your story matters.
Citation Polinsky, Maria, and Olga Kagan. 2007. Heritage languages: In the
'wild' and in the classroom. Language and Linguistics Compass
1(5): 368-395.
Published Version doi:10.1111/j.1749-818x.2007.00022.x
Accessed March 23, 2018 1:43:42 PM EDT
Citable Link http://nrs.harvard.edu/urn-3:HUL.InstRepos:3382973
Terms of Use This article was downloaded from Harvard University's DASH
repository, and is made available under the terms and conditions
applicable to Other Posted Material, as set forth at
http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-
use#LAA
1
DRAFT
REVISED JUNE 8, 2007
Accepted for publication in
Language and Linguistics Compass
Heritage languages: in the "wild" and in the classroom
Maria Polinsky, Harvard, and Olga Kagan, UCLA
1. Introduction
“There is no place for the hyphen in our citizenship. ...We are…not a polyglot
boarding house,” wrote Theodore Roosevelt in “The Square Deal in Americanism”
(Roosevelt 1918). Had Roosevelt lived long enough to meet Jim, class of ‘09 at a
prestigious American university, he would have been pleased. Jim, who is planning to go
to law school, has many talents: he is a member of the debate club, and editor-in-chief of
the school newspaper; he plays the flute, holds numerous winning titles in wrestling, has
been on the dean’s honor list since his freshman year, and is universally liked as a
charming, articulate, caring person. He is an all-American, well-rounded guy. Jim has a
dark secret, though: he has not really spoken to his grandparents since he was five, he
cannot write to them, and he cannot sign his name.
Jim’s grandparents speak no English. Jim’s birth name is Cho Dong-In. Jim is a
Korean American who can barely speak in Korean. In short, he is a hyphenated American
who would have made Teddy Roosevelt happy. In fact, Roosevelt should not have
worried in the first place: about a third of incoming college students in early twenty-first
2
century America are like Jim (Brecht and Ingold 1998). They “speak American” and their
knowledge of the language they were exposed to from birth can range from limited to
non-existent. Thirty years ago these people were called semi-speakers (Dorian 1981), and
they have also been called incomplete acquirers (Montrul 2002, Polinsky 2006),
unbalanced, dominant, or pseudo-bilinguals (Baker and Jones 1998), early bilinguals
(Kim et al. 2006), or speakers of “kitchen language ___” (fill in the blank). The
multiplicity of terms may be, in part, a reflection of the general lack of clarity among
linguists about what such speakers know of their home language, and how best to
characterize them linguistically. About ten years ago, a term that had been used earlier in
Canada (Cummins 2005: 585) to denote this elusive group of speakers made its way to
the U.S: heritage speakers.1
Two conceptions of heritage language have been proposed; we will refer to them
as broad and narrow. The broad conception of heritage language emphasizes possible
links between cultural heritage and linguistic heritage. A definition by Fishman (2001:
81) stresses a “particular family relevance” of a language, and Van Deusen-Scholl (2003:
222) defines those who “have been raised with a strong cultural connection to a particular
language through family interaction” as language learners (not speakers) “with a heritage
motivation”. While “heritage motivation” and “family relevance” are important
impetuses for learning a language, they are not sufficient to charcterize linguistic
1 While it has gained momentum, the term has also encountered opposition from some
researchers and educators who think the term focuses on the past rather than the present
and future. Nevertheless it has been used extensively since the first Heritage Conference
in 1999 (Peyton et al. 2001).
3
knowledge proper, and do not provide operational criteria for identifying heritage
speakers. Culturally motivated learners who learn their heritage language from scratch as
adults are regular second-language speakers, albeit with a different motivation; their
learning trajectory sets them apart from speakers like Jim, described above. For broadly
defined heritage speakers, the heritage language is equivalent to a second language in
terms of linguistic competence, and as a second language it typically begins in the
classroom, in adulthood; for speakers like Jim, their heritage language begins in the
home, and often stops there, too. While we recognize cultural heritage as an important
motivating factor, we would like to underscore the difference between a culturally
motivated L2 learner and a true heritage speaker in the narrow sense of the term, the
sense to which we now turn.
The best known, and most widely used, definition of heritage speakers under a
narrow conception of the term is Valdés’s (2000). She refers to heritage speakers as
“individuals raised in homes where a language other than English is spoken and who are
to some degree bilingual in English and the heritage language.”2 Although the original
definition is English-centered, any other dominant language can be substituted for
English. The crucial criterion is that the heritage language was first in the order of
acquisition but was not completely acquired because of the individual’s switch to another
2 The definition of a heritage speaker in general and for specific languages continues to
be debated. The debate is of particular significance in such languages as Chinese,
Arabic, and languages of India and the Philippines, where speakers of multiple languages
or dialects are seen as heritage speakers of a single standard language taught for
geographic, cultural or other reasons (Mandarin Chinese, Classical Arabic, Hindi, or
Tagalog respectively).
4
dominant language. The other critical component of this definition has to do with
identifying continua of proficiencies, reflecting the tremendous variation in heritage
language observed by researchers.
As often happens with word magic, the arrival of the name ‘heritage speakers’
created a new field. Settling on the term helped researchers identify the research
questions associated with this group of speakers. In this survey article, we will
concentrate on several questions that we consider to be of primary importance:
• What is known about heritage languages and the range of variation among
them?
• What is the structure of a heritage language and what does it tell us about
human language potential in general?
• Can the degree of maintenance and attrition of a particular heritage
language be predicted for a given individual or community? and,
• What pedagogical challenges do heritage languages present?
As the reader will quickly notice, many of the general questions we pose lead us
to new questions, still awaiting answers. It is our hope that this article, which is intended
to serve as much as a promissory note as an assessment of the emerging field, will
stimulate new research seeking to answer the outstanding questions below.
2. Heritage language “in the wild”
2.1. Variation and classification
Anyone who has even a casual acquaintance with heritage language speakers, be that a
language instructor or a second language learner of the language that heritage speakers
learned in the home, will be quick to tell you that all of these speakers are different and
5
seem to know different things, and that many of them sound native-like. We will address
these speakers’ often native-like pronunciation in section 3; in this section, we will
discuss the apparent range of variation found in heritage populations.
The idea that any given language is subject to speaker variation at every possible
level is not new: ask five English speakers to judge the following sentences and you will
get five different answers, from completely unacceptable to just fine:
(1) The birds the linguists they commented on the work of studied are very smart.
(2) Which prizes did you wonder how many people would eventually be awarded?
In heritage languages, however, the range of variation is much greater, and it often leads
to the suggestion that heritage languages are not systematic or, using a milder version of
this claim, do not have any categorical distinctions. We would like to argue that this is
not true. The illusion of endless variation comes from our neglecting to look closer and
recognize groups within the accepted variation range.
Because of their exposure to the home language throughout childhood and, for
some, even into early adulthood, heritage speakers’ strongest suit is generally aural
comprehension. Their speaking abilities fall within a continuum, from rather fluent
speakers, who can sound almost like competent native speakers, to those who can barely
speak the home language. The continuum is reminiscent of what has been proposed in
creole studies. For creole languages, there usually is not one standard variety of the creole
to which most speakers conform, but rather there is a continuum of creole varieties,
ranging from the variety closest to the lexifier (the acrolect) to the least lexifier-like (and
6
often least socially prestigious) variety, known as the basilect (Bailey 1973, 1974,
Bickerton 1973, 1975):
(3) { basilect mesolect acrolect } lexifer
← …….. →
Although the idea of a creole continuum has been subject to numerous
refinements and modifications (see Rickford 1987 for an insightful discussion), the
general conception of a scale which accommodates most of the variation remains quite
solid. This conception allows researchers to capture variation by measuring a particular
variety’s distance from the baseline. In heritage language populations, those heritage
speakers who can be objectively shown to be near-native, maximally close to a competent
(albeit not formally or fully educated) speaker, can be characterized as acrolectal, high-
proficiency speakers. Those who are maximally removed from native attainment and who
show many deviations from the baseline would correspond to basilectal, lowest-
proficiency speakers. Typically, the access of such lowest-proficiency speakers to the
baseline has been restricted to home language (which they often only overhear, cf. Au
and Romo 1997), and often they have been isolated from the community in which their
home language is dominant (we will refer to such communities as the ‘metropoly’).
A common profile of such a basilectal speaker would be that of children growing
up in the U.S. in a family where the home language is not English; such children hear the
home language as spoken by one or both parents, and sometimes by other family
7
members or caretakers. Once they go to preschool, their exposure to the home language
continues to be limited to home communication, and many such speakers very quickly
get into the way of hearing the home language but responding in English (or whatever the
dominant language may be). Additionally they may have never learned to read or write in
the home language, and if the alphabet is non-Latin, this naturally makes matters even
more difficult. By the time they get to college and attend a class in their home language,
they may address their instructor as “Aunt Amy” or “Uncle George”, and use the familiar
form of address or more direct imperatives than are usually expected in formal
contexts—forms that they may have picked up in interaction at home while growing up.
An acrolectal speaker, on the other hand, is someone who has limited deficiencies
in some registers of the baseline language and may even have good awareness of the
standard, in addition to the baseline. Overall, the proposed continuum model for heritage
languages would be as follows:3
(4) { basilect mesolect acrolect } baseline
heritage heritage heritage
← …….. →
The proposed continuum representation allows us to classify the variation found
among heritage speakers in the same way as has been applied to creoles. However, it
requires knowledge of the baseline for comparison and an evaluation metric that can be
3 In this survey we focus mainly on basilect and acrolect, but the recognition of finer
distinctions on the continuum is, of course, important.
8
used to assess a speaker’s distance from this baseline. In what follows, we will take up
these issues in turn.
2.2. Establishing the baseline
In the case of an idealized creole continuum, the baseline language is the lexifier—for
example, English as spoken in Jamaica for Jamaican Creole, “colonial” French for
Seychellois, and so on. But what constitutes the baseline for heritage languages?
The baseline language for a heritage speaker is the language that he or she was
exposed to as a child. A common misconception is that a heritage speaker’s baseline
language is the standard language as promoted by schooling, media, and literature. For
instance, a naïve expectation might be that someone who grew up in Orange County,
California, speaking Vietnamese at home, would know contemporary standard
Vietnamese, which is based on the Northern dialect. In fact, all second-language learning
textbooks on Vietnamese reflect this standard, as do the materials for heritage speakers
used, for example, in the heritage Vietnamese classes currently offered at UCLA or
UCSD.
However, it is the southern dialect of Vietnamese that is spoken by the majority of
Vietnamese speakers in Orange County, and which, therefore, is the baseline for their
children who grow up exposed to it in the home. Thus it is a mistake to expect those
speakers, who have not been through formal schooling in Vietnamese, to have any
knowledge of the standard or Northern dialect. Similar mismatches between the baseline
and standard language occur for heritage Russian speakers in the U.S., most of whom
have had access to the southern varieties of the language (Polinsky 2000) rather than the
standard language associated with Russian literature, schools, and media, and which,
9
until recently, was valiantly maintained by small numbers of early post-communist
émigrés in France or Germany (Zemskaja 2001; see also Polinsky 2003). Likewise, it
would be unrealistic to expect a heritage speaker of Spanish in Florida, where Cuban
Spanish is the norm, to have even rudimentary knowledge of Iberian Spanish or, closer to
home, more standard dialects of Mexican Spanish.
The discrepancy between the baseline and an idealized standard is even more
complicated for languages with well-articulated registers, such as Korean or Japanese.
Korean has six well-differentiated registers, which vary according to the social
distinctions between the speaker and hearer (Sohn 2001: 16, 407 ff.). Each register is
associated with particular phonological patterns and lexical choices (Kim 2001; Choi
2002). Of those six registers expected of any grown-up competent speaker, heritage
speakers of Korean are familiar with one or, at most, two (Kim 2001: 262)—the intimate
register (identified as the -e/-a register in the literature on Korean) and possibly familiar
register (identified as the –ney register). These two registers are used in the home and are
typical of children, but a grown up speaker is of course expected to control the other four
registers as well. However, since heritage speakers of Korean have not been instructed in
more formal registers it would be counterproductive to expect them to control anything
beyond this baseline.
This brief discussion points to a conclusion and an open question. The conclusion
is straightforward: since heritage speakers of language L are typically not exposed to the
language norm through formal schooling, their baseline is the particular variety of L that
such speakers were exposed to in the home, and should not be identified with the
standard language available to fully competent speakers of L. For quite a few languages,
10
including mainstream American English, such a standard is often identified by the
absence of certain features expressly associated with a particular area or social class
(Wolfram and Shilling-Estes 1998: 53-88). Since this standardization is highly
conventional, it takes consistent exposure, for example through schooling, to impose the
standard on an individual; heritage speakers do not have such exposure in their
background.
The outstanding question has to do with the need to establish the baseline for each
particular heritage language. Establishing the baseline requires knowledge of
demographic patterns (who settled where and when) and a good understanding of
dialectal and/or register differentiation in a given language. For some languages, both
ingredients may be readily available to the scholar, but for others work on one or both of
these would need to be done prior to a serious assessment of the heritage language in
question.
This question is also of importance when heritage speakers come to learn their
home language in a formal educational setting. The discrepancies between their spoken
language and the educated norm taught in the classroom may lead some instructors to
discount their knowledge rather than value it. Language instructors tend to look for gaps
in knowledge (which are, of course, plentiful in these cases) rather than assign value to
the rich and varied linguistic backgrounds that these learners bring with them.
This often unintentional difference in approach to second language learners vs.
heritage speakers reminds us of the glass half-full/half-empty metaphor. With second
language learners, every small step toward attainment is celebrated and accepted, and the
necessary gaps in knowledge are viewed as part of the course. In the assessment of
11
heritage speakers, however, the glass is perpetually half-empty: it is hard to ignore what
they don’t know, that they cannot read or write, that they don’t know the standard, and
that they do not speak like true native speakers. While the L2 learner’s achievement, no
matter how limited, is viewed favorably, to some instructors the heritage learner’s
competence is always suspect, and both instructors and curriculum designers tend to take
stock only of their deficiencies. To establish an adequate curriculum for these learners,
we need to adopt the “the glass is half full” approach, i.e. a more optimistic outlook based
on how much heritage speakers already know.
The problem is exacerbated in cases where language instructors are biased against
the language variety that served as the home language. In their study of the attitudes of
members of various Spanish language and literature departments in the U.S. to academic
Spanish as spoken by Spaniards, Mexicans, Latin Americans, and Chicanos, Valdés and
her co-authors (in press) discovered that educators’ views on literacy and prestige dialects
resulted in prejudices favoring certain varieties of academic Spanish and against other
varieties.
2.3. Classifying heritage speakers
Let’s assume that we have established the baseline for a given heritage language. Even
without observing a particular speaker, is it possible to get a quick sense of how close his
or her language is to the baseline? Surprisingly, the answer is a partial yes. Based on
linguistic research on incomplete acquisition and near-native attainment, including the
work done by one of the present authors, several diagnostics can be proposed to establish
a speaker’s proximity to the baseline. One of the most important ones seems to be speech
rate.
12
Speech rate is the word-per-minute output in spontaneous production—as
evaluated, for example, in the description of a picture set.4 The speaker is asked to
describe one set of pictures in her heritage language and another set in English; doing so
provides a standard of comparison for assessing individual variation in speech rate.
Results show that a heritage speaker’s speech rate may be as low as 30% of the average
speech rate for the baseline control, but on the other hand, some heritage speakers come
very close to the baseline rate (Kagan and Friedman 2004).
The relevance of speech rate is attested by a recent study of gender restructuring
in heritage Russian (Polinsky in press-a), which showed that heritage speakers fell into
two distinct groups: (A) those who maintained the baseline three-gender system of noun
classification, with various adjustments, and (B) those who radically reanalyzed the
baseline system as a two-gender system. The latter tendency was strongly correlated with
a lower speech rate. Figure 1 shows the results for average speech rate in monolingual
baseline Russian, monolingual English, and the speech rates in English and Russian for
two heritage speakers, chosen to represent the highest (AR) and lowest (BL) speech rate
in the subject pool.
4 We have been using frog story pictures (Berman and Slobin 1993); these have been
successfully used in many studies involving narratives, with adult and child-language
alike, and as a result, a large body of data, including statistics on several languages, has
been made available by these studies.
13
0
20
40
60
80
100
120
140
160
180
speaker BL speaker AR monolingual
Russian monolingual
English
rate in Russian
rate in English
Figure 1. Average speech rate in heritage Russian, baseline Russian, English spoken by American
Russians, and monolingual English (frog story narrative, words per minute)
Besides adding more evidence to our earlier claim concerning the tremendous variation in
heritage speech, the data in figure 2 also show the correlation between speech rate and
retention of the three-gender system of baseline Russian in heritage speakers: a three-
gender system is correlated with a higher speech rate, while a two-gender system is
correlated with a lower speech rate. The motivation behind this correlation is rather
straightforward. Lower proficiency speakers, who presumably represent the basilect, have
more difficulty in accessing lexical items, which slows down their speech. As we will
show below, knowledge of lexical items and grammatical knowledge are also correlated.
Not surprisingly, problems with lexical access are accompanied by difficulty constructing
phrases and clauses; the spontaneous speech is punctuated by pauses, repetitions, and
false starts. All of this has an immediate effect on the speech rate.
14
0
20
40
60
80
100
120
baseline 3 genders 2 genders
Figure 2. Rate of speech (words/min) in baseline controls, speakers who maintained the three-gender
system, and speakers who switched to a two-gender system for Russian nouns.
Although correlations such as the one shown in figure 2 need to be tested with
other grammatical phenomena and different heritage languages, they offer a promise of
relatively simple diagnostics of language proficiency which can be used both in the lab
and in the classroom.
Speech rate may be a promising method of classification, but it could be difficult
to determine for the lowest-proficiency heritage speakers, who may be reluctant to
produce any language whatsoever. Another diagnostic of proficiency in heritage
languages that may be even more powerful is lexical proficiency. In earlier studies
(Polinsky 1997, 2000, 2006) we have observed a strong correlation between a speaker’s
knowledge of lexical items, measured in terms of a basic word list (about two hundred
items), and the speaker’s control of grammatical phenomena such as agreement, case
marking, aspectual and temporal marking, and use of subordinate clauses, with
grammatical knowledge measured by deviations from the baseline in spontaneous speech
(Polinsky 1997) and, in later studies, by answers to forced-choice judgments (Polinsky
15
2005). The correlation between the levels of grammatical and lexical knowledge was
supported by results for several heritage languages, including Russian, Polish, Armenian,
Korean, and Lithuanian (see also Lee 2001, Godson 2003). This correlation is not
exclusive to heritage language competence; it has also been proposed for early child
language (Bates et al. 1994; Thal et al. 1996, 1997; Fenson et al. 2000). If structural
attrition and lexical proficiency are correlated, lexical proficiency scores, which are
relatively easy to obtain, can serve as a basis for the characterization and ranking of
incomplete learners in terms of the proposed continuum model.
Another good diagnostic of heritage language proficiency relates to the manner
and length of exposure to the baseline. These two characteristics seem interrelated in
ways which may not yet be fully understood. With respect to the manner of exposure, it is
natural to expect that speakers who grew up surrounded by the baseline language in the
metropoly should differ somehow from those who grew up in an immigrant community
in the U.S. or any other country where a different language is dominant. The exposure to
language in the metropoly is inevitably greater than that in immigrant communities, so
one would expect a heritage speaker who spent her first five years of life in Korea to have
an advantage over an American-born Korean heritage speaker. Au and Oh (2005) found
that speaking the majority language before age five seems to put linguistic minority
children at a small, but measurable, risk for poorer heritage language skills during
adolescence. Likewise, the length of exposure to the baseline should also matter, since
the longer the exposure to the baseline, the greater the baseline input for the heritage
speaker.
16
Preliminary results show that the effects of length and manner of exposure are not
easily separable. The only group that seems to have a distinct language advantage are
heritage speakers who grew up in the metropoly and had exposure to the baseline
language at least until the end of the critical period. All other groups are more or less
indistinguishable; crucially, exposure to schooling in the heritage language after
immigration does not seem to have any serious effect, as shown by preliminary studies of
Korean Sunday schools (Kwon and Polinsky 2005) and Chinese Sunday schools (Wang
1996, Xiao 2006, Weger-Guntharp 2006).
Finally, the heritage speaker’s parents have also been found to have an effect on
language maintenance by heritage speakers. Au and Oh (2005) demonstrate that the
language spoken by parents in the home, as well as parents’ attitudes toward the home
language and culture (e.g., instilling ethnic pride, discussing ethnic history and identity,
encouraging children to learn and practice cultural traditions and values) are correlated
with the children’s later abilities in the home language.
3. Heritage speakers’ language knowledge
In this section, we will concentrate mainly on the lowest proficiency, basilectal
speakers, simply because it is their language that shows the most striking deviations from
the baseline. According to the diagnostic measures described above, basilectal speakers
are speakers who grew up with limited or early interrupted exposure to the home
language, and consist of “overhearers” (Au et al. 2002, Oh et al. 2003), who heard the
home language in the background but never responded in it and were not addressed in it
consistently, and bilinguals who switched to the dominant language as early as preschool
(Polinsky 1997, Montrul 2004a). In terms of speech rate, such speakers typically fall in
17
the range of a third to one-half of the speech rate of the baseline (Polinsky in press-a).
Finally, such speakers seem to have a lexical proficiency of about 70%, using the basic
vocabulary technique (Polinsky 1997). Of course, the group of basilectal heritage
speakers may not be as homogenous as we present it here; the impression of homogeneity
may simply be the artifact of our limited knowledge of this group.
Assessing the linguistic knowledge of basilectal speakers is a challenging task—
many of them can barely speak in the heritage language, and persuading them to produce
spontaneous speech, even under the controlled conditions of narrative elicitations, is not
easy (Polinsky 2006; Silva-Corvalan 1994). Aside from this rather torturous spontaneous
production, the bulk of the data on their language knowledge comes from grammatical
judgment tasks; such data are now available for Spanish (Toribio 2001, Montrul 2002,
2004a, b, 2006, , Zapata et al. 2005, Roca 2000), Greek and Italian (Tsimpli et al. 2003,
2004, Sorace 2004), Russian (Polinsky 2005, 2006, in press-a, b), and Korean (Kim et al.
2006).
3.1. Phonetics and phonology
In the sound system of the baseline language, the general impression is that even
basilectal heritage speakers sound native-like, and this gives them a distinct advantage in
re-learning the home language when exposed to it in the classroom, for instance in
college (Au and Romo 1997). This phonetic advantage is often cited as one of the main
reasons for placing heritage and second language learners on different tracks in the
classroom (Peyton et al. 2001). On the other hand, anecdotal evidence that heritage
speakers have a slight ‘accent’ also abounds: competent speakers of the baseline often
18
comment that heritage speakers sound ‘funny’, ‘off’, and not like ‘real’ speakers of the
language.
It is well known that bilingual speakers have different phonetic representations
than monolinguals (Caramazza et al. 1974, Bullock et al. 2004, Sundara et al. 2006), so
the presence of a heritage ‘accent’ should not come as a total surprise. However, there is
emerging evidence that heritage speakers differ from more balanced bilinguals as well.
For example, an instrumental study of vocalism in heritage Western Armenian (Godson
2003, 2004) showed that heritage speakers differed from both Armenian-dominant
bilinguals and monolingual baseline speakers. The difference could not be attributed
solely to transfer from English: English seemed to play a role only for those vowels that
are already close to English (namely, /i/, /e/). The heritage speakers’ other vowels, and
particularly /o/ and /u/, were quite distinct from their English counterparts and from the
respective vowels in the bilingual system. While the reasons for such differentiation
remain unclear, Godson’s results clearly show that the heritage accent is a measurable
reality. Studies of both consonantal and vocalic systems in other heritage languages are
desirable and may reveal further regularities.
While instrumental studies targeting the phonetics of heritage speech are badly
needed, virtually nothing is known about the nature of phonological representations in
heritage speakers. The lack of studies is due, in part, to the fact that we are dealing with
an emerging field, but is also likely due to the fact that heritage speakers generally sound
so native-like—one could easily imagine that there would be no differences in
phonological representations between the heritage language and the baseline, although
that remains to be shown.
19
3.2. Morphological categories
Turning to morphology, basilectal heritage speakers show an across-the-board
over-regularization in morphological paradigms, with the elimination of irregular and
infrequent forms (see Seliger and Vago 1991 for a number of languages; Silva-Corvalan
1994; Anderson 2001; Montrul 2002, 2004a for Spanish; Halmari 1998, 2005 for Finnish;
Levine 2000 for Yiddish; Polinsky 2000, 2006, in press-a, c for Russian; Caruso 2004 for
Italian; Choi 2002 for Korean; Dutkova 1998 for Czech; Cozens 2003 for Polish).
Complicating the picture somewhat is the fact that the bulk of morphological leveling and
over-generalization that has been documented occurs in heritage languages with
interference from English, which itself does not show rich morphology. The few studies
of incomplete acquisition under the dominance of other languages, however, show that
over-regularization happens nevertheless, thus suggesting that transfer from the dominant
language is not the complete answer (cf. Backus 1996, 1999 for the incomplete
acquisition of Turkish in Holland, or Leisio 2001 for the incomplete acquisition of
Russian in Finland).5 Over-regularization is particularly apparent in verb forms, simply
because most languages have a larger variety of verb paradigms to start with. However, it
is also present in nominal and pronominal declensions and other morphological
subsystems.
Two observations are in order here. First, while formal over-regularization is
visible and often occupies the center of researchers’ attention, it is also important not to
5 Superficially, it may seem that there are many more studies of what may be incomplete
acquisition in bilingual communities, but such studies often avoid the question of change
in the “home” language and focus on such phenomena as code-switching.
20
lose sight of over-regularization in interpretation. The underlying principle of such over-
regularization is the elimination of polysemy. For instance, Montrul (2002) shows that
heritage Spanish speakers do not use achievement verbs in the imperfect. Achievement
verbs typically describe an event that has an endpoint; the imperfect tense is often,
although not always, associated with the absence of such an end point. It seems that
heritage speakers may over-generalize the semantics of the imperfect to the extent that
they never use it to indicate completed action. Language instructors are familiar with
heritage speakers questioning whether imperfect verbs can even be used in the past tense
because the action they express does not have an end point. Another observation
concerning the reanalysis of tense and aspect in heritage speakers has to do with the
avoidance of the past tense with stative verbs (examples of stative verbs include verbs
expressing feelings: ‘like’, ‘hate’, ‘love’, ‘be sad’, etc.) (Montrul 2002 for Spanish;
Polinsky 2006 for Russian). The past tense can easily be generalized as locating an event
in time prior to the reference point; meanwhile, stative verbs are often associated with the
absence of temporal limits, with the same result as described above for imperfect verbs.
In addition to overgeneralizing in both form and meaning, heritage speakers are
also extremely good at maintaining fossilized forms of high-frequency items—for
instance, count forms (even in the absence of the full-fledged nominal paradigm), frozen
adpositional (pre- and postpositional) phrases referring to time or location (‘on Monday’,
‘at home’, ‘with me’), or occasional polite imperatives. Such fossilized forms or ‘chunks’
may make it seem that a heritage speaker controls more morphology than she actually
knows, and they certainly add to the impression of tremendous variation among heritage
speakers. If our knowledge of heritage languages remains confined to heritage speakers’
21
production, especially spontaneous production, it is inevitable that such fossilizations will
cloud the issues in this way. Only systematic perception experiments will allow us to
separate fossilized chunks from the speakers’ more productive knowledge of grammar.
The presence of fossilized chunks in heritage speech raises a more general
question regarding the role of frequency in heritage language grammar. What if all that
heritage speakers know is just a combination of high frequency items that were present in
the input? The answer to this question is important not only for our understanding of
heritage languages, but also for the ongoing debate on the role played by positive and
negative evidence in language acquisition. Few people, if any, doubt that frequency of
input is important to acquisition; however, it is not clear whether frequency plays a direct
or more mediated role in language learning. In our study of heritage speakers’ knowledge
of lexical classes (parts of speech), we addressed this question from the standpoint of the
interaction between frequency and structured grammar (Polinsky 2005). Heritage
speakers showed a strong differentiation of lexical items of identical frequency based on
their lexical class: they had a much better control of verbs over nouns and adjectives of
the same frequency. Within each lexical class, more frequent items were accessed faster
than lower frequency items, but the differentiation across lexical classes was much more
pronounced. While studies such as this are awaiting replication, we can hypothesize that
frequency cannot be all there is to incomplete acquisition.
3.3. Syntactic structures
The syntactic characteristics of basilectal heritage language are still poorly understood
but some generalizations have started to emerge, and we will focus on those here. As in
the discussion of other language subsystems, we would like to start with methodological
22
observations. Generally, linguistic research in syntax relies heavily on speakers’
intuitions as elicited in the form of grammaticality judgments. The reliance on intuitions
has received its fair share of criticism from linguists working on uninterrupted acquisition
and completely acquired languages; with respect to heritage language, however,
judgments are particularly hard to elicit and assess. It is quite striking that low-
proficiency heritage speakers perform at chance on staple grammaticality judgments that
linguists are so good at using, for example, with respect to so-called ‘binding’
phenomena, that is, grammatical constraints on the distribution of reciprocals, reflexives,
pronouns and proper names6 (see Polinsky 1997, 2006 for data on heritage Russian).
We are not sure what may motivate low-proficiency speakers’ reluctance to
produce grammaticality judgments, but at this stage we would like to offer a cautionary
note regarding the actual use of forced choice with low-proficiency heritage speakers.
Since heritage speakers are often very limited in what they can produce spontaneously,
the potential for observation of their naturally occurring speech is also limited. At the
level of methodology, this suggests that we should rely on the standard language tasks
used with populations whose speaking ability is limited (e.g. child speakers, aphasics);
6 These constraints, which together constitute the Binding Theory in generative grammar,
are responsible for the failure of the pronoun and the proper name to refer to the same
person in sentences like ‘John likes him’; on the other hand, when ‘him’ is replaced by a
reflexive, ‘John’ and ‘himself’ must be coreferential: ‘John likes himself’. The robust
nature of competent speakers’ judgments about sentences like these means that the
Binding Theory is a useful diagnostic tool for a whole host of syntactic phenomena; that
heritage speakers should perform no better than chance on these judgment tasks therefore
presents a new challenge to the traditional syntactician.
23
such tasks include picture naming, picture-utterance associations, or simple cross-modal
priming. In our experience, this methodology has been most effective in establishing
what heritage speakers actually know, but it is, of course, painstaking and slow.
In the rest of this subsection, we will focus on several recurrent features of
heritage language grammars. Given the simplification of morphology described above, it
is not surprising that heritage languages show a much smaller range of morphological
case distinctions than the baseline; this has been observed for a wide range of heritage
languages (Seliger and Vago 1991; Polinsky 1996; Halmari 1998). One of the striking
properties of case reanalysis in heritage systems is the use of an unmarked case
(nominative, absolutive) as the generalized case, including as the case used with
adpositions. For instance, in heritage Kabardian, complements of postpositions appear in
the marked (definite) absolutive case (5a), while in the baseline postpositions require the
generalized oblique (5b):7
(5) a. bostexə-r pajə
dress-ABS.DEF for
b. bosteχe-m pajə
dress-OBL for
‘for the dress’
7 Note also pronunciation differences between the heritage form and the baseline form.
24
Another tendency, especially with baseline languages that have an articulated
system of oblique cases, is for the heritage language to have one generalized oblique
case. In heritage Russian, for instance, in the plural, this case is the prepositional in –ax,
which replaces the other oblique cases—compare the examples in (6). Intriguingly, a
similar overgeneralization of the prepositional case is observed in Russian child language
(Gvozdev 1961: 334) and in non-standard dialects of Russian.
(6) Heritage vs. baseline Russian: oblique case marking in the plural
Heritage Baseline
a. bez rukav-ax bez rukav-ov ‘without sleeves’
without sleeve-PRP.PL without sleeve-GEN.PL
b. mnogo stakan-ax mnogo stakan-ov ‘many glasses’
many glass-PRP.PL many glass-GEN.PL
c. za skamejk-ax za skamejk-ami ‘behind the benches’
behind bench-PRP.PL behind bench-INSTR.PL
d. dlja detj-ax dlja det-ej ‘for children’
For children-PRP.PL for children-DAT.PL
In the baseline language, case encodes relations between the case-assigning head
and its dependent; once morphological marking of case is reduced or eliminated, these
relations between the assigner and its dependent must be expressed in some other way.
The potential compensatory strategies for expressing these relations after the loss of
morphological case include agreement and word order. Since agreement typically
requires sophisticated morphology, which heritage language systems do not have, fixed
word order thus becomes a staple feature of heritage language grammars. As a result, in
languages where discontinuous expressions are found, this feature is lost in the heritage
25
variety. For instance, in Korean it is common for a quantifier (a word like ‘all’ or ‘many’)
to appear discontinuously, away from its associated noun phrase; this phenomenon is
known as ‘quantifier float’ (Gerdts 1987, Miyagawa 2006).8 Yet heritage Korean
speakers invariably produce the quantifier adjacent to the noun, and if asked to judge
expressions with floated quantifiers they generally hesitate or even find them
ungrammatical.
In addition, flexible word order languages lose their general ability to scramble
(the ability of lexical items to move from their base positions, resulting in apparent
flexibility of ordering with respect to subject, verb and object); for instance, another
feature of heritage Korean is lack of scrambling from SOV to OSV word order, a feature
observed in the baseline, hence indicating the freezing of surface word order (Kwon and
Polinsky 2005). Notably, heritage speakers of those baseline languages that can invert
subject and verb in certain constructions (such as Spanish, Finnish, and Russian) typically
do not show the expected verb-subject order (Silva-Corvalan 1994; Sanchez 1983;
Halmari 1997; Isurin and Ivanova-Sullivan 2007).
While it would be tempting to motivate the rigid word order found in heritage
grammars by universal tendencies in language encoding, we do not think that the time for
such a conclusion has arrived. As mentioned above, most of our data on heritage
languages come from speakers whose dominant language is English. With its own
shallow morphology and rather rigid word order, it may simply be that English acts as a
8 Quantifier float is present in English as well, for example in (ii), where each ‘floats
away’ from its host noun children:
(i) The clown gave each child two balloons
(ii) The clown gave the children two balloons each
26
source of transfer (cf. Ionin 2007). If so, data on heritage languages spoken in
communities where another language is dominant are needed to decide between the
universalist explanation and the transfer explanation of word order effects observed in
low-proficiency heritage speakers.
In addition to the more surface-level phenomena related to case marking and
word order, heritage speakers also have difficulty maintaining syntactic dependencies
pertaining to a more abstract level of syntactic representation, what was traditionally
termed ‘Deep Structure’. An example might be so-called ‘long-distance dependencies’9
These configurations involve a relationship between two syntactic positions across a
stretch of lexical material, where the lexical item in one of these positions (a pronoun, for
example) is not on its own sufficient to determine its reference (say, the particular person
that it picks out), and is instead dependent upon the content of a noun phrase elsewhere in
the sentence. In such a case, the more fully articulated noun phrase is termed the
‘antecedent’ of the underspecified item, and the two terms are conventionally marked
with the same subscript index to indicate that they refer to the same individual. For
example, a lexically specified noun phrase can serve as the antecedent of a pronoun
(including a null pronoun) (7a), a reflexive anaphor (7b), or even an element that is not
pronounced at all, but is nevertheless present in syntax as a ‘gap’ or null element (7c, d).
(7) a. A reporter asked the senatori what shei hoped to accomplish with the new bill.
b. A reporter asked the senatori about herselfi.
9 This seems to apply both in production and comprehension, although data on
comprehension are extremely scarce.
27
c. A reporter asked the senatori at the press conference ___i to elaborate on the new
proposal.
d. A reporter interviewed the senatori [that everyone admired ___i]
Without going too much into detail, we would like to make a couple observations
here. First, heritage speakers have significant difficulties producing null elements. For
structures such as (7c, d), a typical strategy might involve the use of a pronoun instead of
a gap. Given that verbal morphology is not fully represented in heritage speech, this often
leads to an impression that heritage speakers have difficulty with embeddings. In
addition, the absence of null elements creates an impression of extreme redundancy, with
pronouns commonly appearing where the baseline language does not show any. As an
illustration, consider the following excerpt from a frog story (heritage Russian speaker,
23 y.o., interrupted at about 5):10
(8) mal’čik # on imel sobaka i ljaguška.
boy 3SG had dog.DC and frog.DC
# on ljubit ego ljaguška
3sg likes his frog.DC
sobaka # on tože ljubit # ljaguška.
dog 3SG also likes frog.DC
mal’čik # on spat’. on spit.
boy 3SG sleep.INF 3SG sleeps
sobaka tože spal, ljaguška vyxodit iz # jar # i on uxodil.
10 Special symbols: code switching is italicized, pauses are indicated by #; DC—default
case, INF—infinitive.
28
dog also slept frog comes_out of and 3SG was_leaving
on uxodil iz ix dom sovsem
3SG was_leaving from [their house].DC completely
The heritage speaker’s use of pronouns therefore amounts to the following non-native
sounding sentences in English (and makes the same impression of non-nativeness in
Russian):
‘The boy HE had a dog and a frog. The dog HE also likes the frog. The boy HE
slept.”
A more natural translation of the intended story (and a corresponding narration in
Russian that could be expected of a full Russian speaker) would be: ‘The boy had a dog
and a frog. The boy and the dog liked their frog. Once when the boy and the dog were
asleep, the frog got out of the jar where it lived and left the house for good.’
Somewhat related to the avoidance of null elements is the difficulty heritage
speakers experience with reflexivization. While baseline Russian has standard possessive
reflexives, the heritage speaker in (8) does not use them in referring to the frog or the
house (that is, the speaker says ‘his frog’, while a full Russian speaker would be expected
to use a special pronoun that means ‘one’s own’, something like ‘self’s frog’).
Spontaneous production and judgment tasks have confirmed this failure to use possessive
reflexives, but systematic experimental work is needed to determine the extent to which
heritage speakers have difficulty with long-distance dependencies and why.
29
4. Heritage learners in the classroom
While we focused above on heritage speakers for the purposes of linguistic research, for
pedagogical purposes we will focus on heritage learners, namely heritage speakers who
wish to regain, maintain or improve their home language through classroom instruction.
The challenge for the teaching profession is to find pedagogical solutions relating
to heritage learners based on current linguistic research, a task that clearly has additional
practical significance. Otherwise, many heritage learners will continue falling through
the proverbial cracks and missing their chance to re-gain proficiency in the home
language, as found in a case study described by Wiley (in press). Wiley analyzes the case
of a speaker raised in Taiwan until the age of five, when he was brought to the United
States and rapidly acquired English. He spoke both Taiwanese and some Mandarin,
although with a Taiwanese accent, and when he enrolled in Chinese classes while in
college, he failed to advance his knowledge of Mandarin because no course took into
account his background and his Taiwanese accent was looked down upon by his
instructors. Wiley writes, “Had the course been designed to take into account the
background and needs of ‘dialect’ speakers, his regional ‘Taiwanese’ accent in Mandarin
might not have been treated in a stigmatizing way”, and the heritage Chinese speaker
would not despair at (re)learning the language. We can substitute another language for
Chinese, but the story of frustrated language study will remain the same for very many
heritage speakers.
As seen in Wiley’s case study, the differences between a heritage learner’s home
language and the form of the language taught in the classroom are often significant, and
can severely hinder a heritage learner’s attempts to improve his knowledge of the
30
language in a classroom setting. While it is not realistic to expect each home language to
be taught separately and with its own course materials, by identifying such mismatches
between home language and the standard or other variety of the language taught in the
language course, instructors can better accommodate heritage learners and help them
achieve their desired language goals.
Among the other challenges of heritage language classroom instruction are the
assessment of heritage learners’ abilities and curriculum development. We will offer
preliminary suggestions with respect to these challenges below.
4.1. Correlating the heritage language continuum with existing
classifications
To better assist language learning specialists to utilize the theoretical results
described above, we will now correlate the groupings found by researchers of two well-
studied heritage language groups—heritage Spanish and heritage Russian speakers—to
the heritage language-continuum model proposed herein. In addition, we will propose
correlations between the groupings proposed herein and the ratings of L2 abilities used in
academia (American Council on the Teaching of Foreign Languages, ACTFL, 1999)11
and by U.S. government agencies (Interagency Round Table, ILR).12
In her study of heritage Spanish speakers, Valdés (2001) distinguishes between
three groups of heritage language students. First, she identifies a group of English-
dominant, third- or fourth-generation U.S.-born Hispanic students who have limited
11http://www.sil.org/lingualinks/languagelearning/OtherResources/ACTFLProficiencyGui
delines/contents.htm
12 http://www.utm.edu/staff/globeg/ilrhome.shtml.
31
speaking skills and no literacy in Spanish, all of whom are the least proficient in heritage
Spanish; this group consists of basilectal speakers. The second group she identifies
consists of first- or second-generation bilinguals who possess different degrees of
proficiency in English and Spanish; these students can be expected to display
characteristics of basilectal and mesolectal speakers. Her third group consists of recent
immigrants to the United States from a Spanish-dominant metropoly and who are
Spanish-dominant; we hypothesize that most members of this group would correspond to
acrolectal speakers.
Three groups have also been identified among heritage Russian learners (Kagan
and Dillon 2001, Kagan 2005). Group 1, the most proficient group, comprises students
who graduated from high school in Russia or a former republic of the Soviet Union, and
who have a fully developed grammatical system, a native range of vocabulary, and inside
familiarity with Russian culture. Group 2 consists of students who attended school in a
Russian-speaking country for five-to-seven years (an approximate equivalent of
American junior high school) and whose language development in their teenage years
was interrupted by emigration. Heritage learners from Groups 1 and 2 are the closest to
native speakers in grammatical accuracy and breadth of vocabulary; their speech rate is
typically quite high, and their lexical knowledge is around 90-95% on the basic
vocabulary test (Polinsky 1997, 2006). These learners correspond to acrolectal speakers
discussed above. Group 3, the members of which can be characterized as ‘incomplete
acquirers’ and ‘forgetters’ (Polinsky 2000), consists of two main subgroups: (a) students
who attended elementary school in a Russian-speaking country, and (b) students who
emigrated as preschoolers or were born in the United States to Russian-speaking parents,
32
and who have been educated entirely in English. These speakers would be classified as
basilectal speakers under the system we propose.
Although designed for the assessment of foreign language learners rather than
heritage language learners, the ACTFL and ILR ratings may also be roughly correlated to
the model of heritage language proposed herein. The ACTFL ratings distinguish between
Novice, Intermediate, Advanced and Superior competencies, with Novice, Intermediate
and Advanced competencies divided into three sublevels (low, mid, and high), and the
ILR system uses a numerical scale from “0”, for speakers with no knowledge of the
language, to “5”, for speakers with the ability of educated native speakers. Comparing
these systems to the abilities of basilectal, mesolectal, and acrolectal learners, we have
arrived at the general correlations shown in Table 1.
Heritage language model ACTFL ILR
Basilectal speakers Intermediate Low/Mid 1: elementary proficiency
Intermediate High 1+: elementary proficiency plus Mesolectal speakers
Advanced Low/Mid 2: limited working proficiency
Acrolectal speakers Advanced High 2+: limited working proficiency,
plus
Acrolectal speakers
native speakers
Superior 3-5: general proficiency, advanced
proficiency, functionally native
proficiency
33
Table 1. Correspondences between heritage language continuum model and ACTFL and ILR ratings (after
NVTC 2006)
4.2. Assessment challenges
Accurate assessment of a heritage learner’s language ability is essential to placing the
learner in a suitable language course, measuring the learner’s language achievement, and
predicting the learner’s achievable proficiency in the language.
Placement tests are of particular interest to language instructors who either have
separate tracks for foreign language learners and heritage language learners, or use a
differentiated approach in a mixed class (Carreira 2004). However, since foreign
language placement tests are typically textbook-based, such tests are not suitable for
placing heritage learners, whose language knowledge was not acquired from a textbook.
As a result, placement of heritage learners according to the results of such placement tests
often results in frustration and dissatisfaction for learner and instructor alike.
To better determine the appropriate placement of heritage learners, a three-
component testing procedure has previously been suggested, consisting of (i) an oral test
loosely based upon the ACTFL oral proficiency interview; (ii) a short essay (if the learner
is literate in the heritage language); and (iii) a biographic questionnaire (Kagan 2005).
Each component may provide information on a separate aspect of the heritage learners’
knowledge; the oral interview provides information on spoken and aural proficiencies,
the essay demonstrates the degree of literacy, and the biographic questionnaire provides
information about language use in the home and the amount of input and output. The
results of self-evaluations of language competencies may also be useful as an assessment
34
device. Typically, heritage learners are best at listening, followed by speaking, reading
and writing. Figure 3 shows the preliminary results of a survey currently being
conducted at UCLA in which heritage learners are asked to assess their language
ability.13
Figure 3. Self-reporting of language ability in the heritage language (data from an ongoing study conducted
at UCLA).
As practice shows, the three-component placement procedure is viable, but a
more sophisticated testing procedure based on linguistic research may also be suggested.
As discussed above, the correlation between a heritage speaker’s levels of grammatical
and lexical knowledge indicates that a lexical proficiency test may be a reliable
13 The data has been obtained from an ongoing survey of heritage learners conducted by
the Title VI National Heritage Language Resource Center. Funding for the study
provided by the US Department of Education Title VI grant # P229A060008.
Please rate your heritage language abilities
0
50
100
150
200
250
300
350
400
450
none
low
intermediate
advanced
native-like
Listening
Speaking
Reading
Writing
35
diagnostic for determining a heritage learner’s proficiency, and one that could be
performed quickly and without elaborate testing procedures. Combining such a test with
an oral interview and a biographic questionnaire may constitute an improvement to the
three-component evaluation described above.
4.3. Developing a heritage-specific curriculum
Although a substantial amount of work has been done on the determinants of
successful second-language acquisition, much less is known about the particular factors
that determine a heritage learner’s success or failure in her attempts to study her home
language in the classroom. Most existing curricula for heritage learners are adapted from
those designed for foreign language teaching, by increasing either the pace or the amount
of material to be covered. However, to best meet the needs of heritage learners, a
differentiated, learner-centered approach may be required (Carreira 2004).
Valdés (1997), for example, sees the goals of instruction for Spanish heritage
speakers as language maintenance, expansion of the bilingual range, development of
literacy skills, and acquisition of a prestige variety, standard dialect or literary norm.
Similar goals are typically set by other heritage language programs as well. However,
unlike instruction in foreign languages, heritage language instruction is not yet viewed as
a multi-year sequence with clearly defined goals, a separate curriculum, and measurable
outcomes. There is a need for a matrix that could serve as the foundation for a program
for heritage language learners (Kagan and Dillon 2003: 100). But as Valdés aptly notes,
the teaching of heritage language learners continues to lack a theoretical underpinning
because “theories that can directly support their teaching” have not yet been developed
(1995: 308). Pedagogical theory relating to the instruction of heritage language learners
36
cannot be formulated without language-specific research, like that referred to in this
paper, to identify those features of heritage languages that tend to be particularly
fossilized or underdeveloped. Heritage language ability-specific research is also
necessary, because, as we have attempted to demonstrate, the grammars of low- and high-
proficiency heritage speakers differ greatly, as do the corresponding pedagogical needs.
At the same time, language educators need to engage in a longitudinal study of
pedagogical methodologies and outcomes to determine what works and what does not
work in the classroom instruction of heritage language speakers. However, the
evaluation of different methodologies requires measurable outcomes. Such outcomes can
only be meaningful if the goals of heritage language instruction have been developed and
accepted by heritage language professionals and learners alike. Developing such goals
based on solid research to determine what can and cannot be achieved within the limits of
a reasonable amount of instruction is the greatest challenge we are facing at the moment.
5. Conclusions
Some of the excitement of heritage language study lies in its novelty—new things are
always exciting. It is also exciting because it brings together several related fields of
inquiry that, unfortunately, are not in the habit of talking to each other: theoretical
linguistics, with its emphasis on universal principles of language structure; experimental
linguistics, especially the study of comprehension, which stands to gain a lot from
working with barely responsive, but readily available, populations; acquisition, which can
now compare happy and arrested development, and language teaching and pedagogy,
which needs to revisit some of the old tried-and-not-so-true L2 methodologies.
37
We have shown that heritage languages are still an unchartered territory, both
linguistically and pedagogically, but we would like to end this paper on an optimistic
note, underscoring how much these languages have to offer linguistic theory. A parallel
that immediately comes to mind is with the study of creoles. Some forty years ago,
creoles were the domain of specific language study or sociolinguistics, and no respectable
linguist worth his stripes would go near them. As soon as linguists recognized that creole
phenomena speak directly to Plato’s problem in language, creoles gained visibility in
linguistic theorizing. Heritage languages add yet another piece to the puzzle of how a
grammar can be acquired under minimal input. Although we are only beginning to
understand how heritage languages are structured, the emerging patterns point to
interesting differences between complete and incomplete first language acquisition, as
well as second language acquisition by heritage speakers and foreign language learners
(Montrul 2004b, Ionin 2007). The emerging evidence shows that grown-up heritage
speakers do not simply hold on to fossilized, frozen grammars from their childhood
(Polinsky in press-a). Instead, the grammar undergoes a reanalysis, but what drives this
reanalysis? Answering this question may help us come closer to solving the puzzle of
language learning.
Heritage languages also hold another attraction. Since the 1990s, linguists have
become increasingly aware that the study of language should no longer be solely the
prerogative of introspective investigation. Instead, language is something that can be
measured using standard experimental methods, and modeled on the basis of rigorously
established data. Nowhere is this paradigm shift more welcome than in studies of those
who can barely produce language: children, aphasics, and aging populations. Several
38
times in this survey article we have mentioned the fact that heritage speakers often have
problems with spontaneous production, which calls for the development of new methods
to discover their language knowledge. Unlike children or patients with speech disorders,
heritage speakers are easy to find, they are motivated and cooperative, and they may even
become active participants, not just experimental subjects, in the study of the extent of
their linguistic knowledge.
Finally, not only do heritage language speakers present us with a wonderful
linguistic challenge, they are also an untapped national resource. We have surveyed
studies showing that heritage speakers have an advantage over L2 learners in re-learning
their home language. This advantage may not always be visible and is more apparent in
phonology and lexicon than in morphosyntax or discourse structure, but it is a reality
nonetheless. The globalization of our economy and recent political turmoils have brought
new urgency to the need for corporate and government employees fluent in the languages
and customs of the countries with which our nation has political and economic ties. The
knowledge possessed by heritage learners places them years ahead of anyone studying
the language from scratch. This it only makes sense to give them the opportunity to
develop further the abilities they already have. The introduction of heritage language
courses reflects the acknowledgment that heritage language speakers are a very special
group, and in some sense a severely underutilized national resource; with proper
instruction, they are much more likely than any second language learner to achieve near-
native linguistic and socio-cultural fluency. That said, we need to be able to formulate
what the ‘proper’ instruction is to be. This is an exciting and daunting task for language
educators.
39
REFERENCES
ACTFL (American Council on the Teaching of Foreign Languages): 1999, Proficiency
guidelines—speaking. Foreign Language Annals 33(1). 13-18.
Anderson, Raquel T. 2001. Lexical Morphology and Verb Use in Child First Language
Loss: A Preliminary Case Study Investigation. International Journal of Bilingualism 5.
377-401.
Au, Terry, and Janet Oh. 2005. Korean as a heritage language. Handbook of East Asian
Psycholinguistics, Part III: Korean Psycholinguistics, Ping Li general editor. Cambridge:
Cambridge University Press.
Au, Terry and Laura Romo. 1997. Does childhood language experience help adult
learners? The Cognitive Processing of Chinese and related Asian Languages, ed. by
Chen, 417-441. Beijing: Chinese University Press.
Au, Terry. K., Leah M.Knightly, Sun-Ah Jun, and Janet S. Oh. 2002. Overhearing a
language during childhood. Psychological Science 13. 238-24.
Backus, Ad. 1996. Two in one: bilingual speech of Turkish immigrants in the
Netherlands. Tilburg: University of Tilburg.
Backus, Ad. 1999. Mixed Native Languages: A Challenge to the Monolithic View of
Language. Topics in Language Disorders 19(4). 11-22.
Bailey, Charles-James N. 1973. Variation and linguistic theory. Arlington, VA: Center
for Applied Linguistics.
Bailey, Charles-James N. 1974. Some suggestions for greater consensus in creole
terminology. Pidgins and Creoles: Current Trends and Prospects, ed. by David DeCamp
and Ian Hancock, 88-91. Washington, DC: Georgetown University Press.
Baker, Colin, and Sylvia Prys Jones. 1998. Encyclopedia of Bilingualism and Bilingual
Education. Philadelphia: Clevedon Multilingual Matters.
Bates, Elizabeth, Virginia Marchman, Donna Thal, Larry Fenson, Philip Dale, J. Steven
Reznick, Judy Reilly, and Jeff Hartung. 1994. Developmental and stylistic variation in
the composition of early vocabulary. Journal of Child Language 21. 85-124.
Bickerton, Derek. 1973. The nature of a creole continuum. Language 49. 640-69.
Bickerton, Derek. 1975. Dynamics of a creole continuum. Cambridge: Cambridge
University Press.
40
Berman, Ruth A. and Dan I. Slobin. 1994. Relating Events in Narrative: A
Crosslinguistic Developmental Study. Hillsdale, NJ: Lawrence Erlbaum Associates.
Brecht, Richard D., and Catherine W. Ingold. 1998. Tapping a national resource:
Heritage learners in the United States, ERIC Digest EDO-FL-98-12, ERIC Clearinghouse
on Languages and Linguistics, Washington D.C.,
www.cal.org/ericcll/digest/brecht01.html
Bullock, Barbara, Almeida Jacqueline Toribio, Kristopher Davis, and Christopher Botero.
2004. Phonetic Convergence in Bilingual Puerto Rican Spanish. Proceedings of the West
Coast Conference on Formal Linguistics 23. 113-125.
Caramazza, Alfonso, G.H. Yeni-Komshian, and Edgar Zurif. 1974. Bilingual switching:
The phonological level. Canadian Journal of Psychology 28. 310-318.
Carreira, Maria. 2004. Seeking Explanatory Adequacy: A Dual Approach to
Understanding the Term “Heritage Language Learner”. Heritage Language Journal 2 (1).
http://www.heritagelanguages.org
Caruso, Marinella. 2004. Attrition in the Verb System of Italian in Australia. Australian
Review of Applied Linguistics, supplement 18. 9-24.
Choi, Hye-Won. 2002. Pragmatically-driven morphological change: Paradigm leveling in
irregular verbs of American Heritage Korean. Paper presented at the UCSD Linguistics
Colloquium, February 2002.
Cozens, Sarah. 2003. Heritage Learners in the Polish Language Classroom. Australian
Slavonic and East European Studies 17 (1-2). 27-45.
Cummins, James. 2005. A proposal for action: Strategies for recognizing heritage
language competence as a learning resource within the mainstream classroom.
The Modern Language Journal 89(4). 585-592.
Dorian, Nancy. 1981. Language death: the life cycle of a Scottish Gaelic dialect.
Philadelphia, PA: University of Pennsylvania Press.
Dutkova, Lida. 1998. Texas Czech: An ethnolinguistic study. University of Arizona:
Ph.D. dissertation.
Fenson, Larry, Elizabeth Bates, Philip Dale, Judith Goodman, J.Steven Reznick, and
Donna Thal. 2000. Measuring variability in early child language: Don’t shoot the
messenger. Comment on Feldman et al. Child Development 71. 323-328.
Fishman, Joshua .A. 2001. 300-plus years of heritage language education in the United
States. Heritage Languages in America: Preserving a National Resource, ed. by J.K.
41
Peyton, D.A. Ranard, and S. McGinnis, 81-98. Washington, D.C., and Delta Systems,
McHenry, IL: Center for Applied Linguistics.
Gerdts, Donna. 1987. Surface Case and Acceptable Relations in Korean: Evidence from
Quantifier Float. Studies in Language 11. 181-197.
Godson, Linda. 2003. Phonetics of Language Attrition: Vowel production and
articulatory setting in the speech of Western Armenian heritage speakers.
UCSD: Ph.D. dissertation.
Godson, Linda. 2004. Vowel production in the speech of Western Armenian heritage
speakers. Heritage Language Journal 2. http://www.heritagelanguages.org.
Gvozdev, Aleksandr N. 1961. Voprosy izučenija detskoj reči. [Issues in the study of child
language] Moscow: Izd. Akad. ped. nauk.
Halmari, Helena. 1997. Government and codeswitching: Explaining American Finnish.
Amsterdam: John Benjamins.
Halmari, Helena. 1998. Case-assignment and adverbials in Finnish-English bilingual
sentences. Language contact, variation, and change, ed. by Jussi Niemi, Terence Odlin,
and Janne Heikkinen. 98-110. Joensuu, Finland: University of Joensuu.
Halmari, Helena. 2005. "I'm forgetting both": L1 maintenance and codeswitching in
Finnish-English language contact. International Journal of Bilingualism 9. 397-433.
Ionin, Tania. in press. DST vs. UG: can DST account for purely linguistic phenomena?
Isurin, Ludimila, and Tanya Ivanova-Sullivan. in press. Lost in between. Heritage
Language Journal. http://www.heritagelanguages.org
Kagan, Olga, and Kathleen Dillon. 2001. A new perspective on teaching Russian: Focus
on the heritage learner. Slavic and East European Journal 45. 507-518. Reprinted in
Heritage Language Journal 1, www.heritagelanguages.org.
Kagan, Olga, and Kathleen Dillon. 2003. Heritage speakers' potential for high-level
language proficiency. Advanced Foreign Language Learning: A Challenge to College
Programs, ed. by Heidi Byrnes and Hiram H. Maxim, 99-112. Boston, MA:
Heinle/Thomson.
Kagan, Olga, and Debra Friedman. 2004. Using the OPI to place heritage speakers of
Russian. Foreign Language Annals 36(4). 536-545.
Kagan, Olga. 2005. In support of a proficiency-based definition of heritage language
learners: The case of Russian. International Journal of Bilingual Education 8. 213-21.
42
Kim, Hi-Sun Helen. 2001. Issues of heritage learners in Korean language classes.. The
Korean Language in America: Volume 6. Papers from the Annual Conference and
Teacher Training Workshop on the Teaching of Korean Language, Culture, and
Literature, ed. by Jungho Ree (6th, Manoa, Hawaii, August 2-5, 2001). 257-274.
Kim, Ji-Hye, Silvina Montrul, and James Yoon. 2006. On the Logophoric Long-Distance
Binding Interpretation of the Korean Local Anaphor 'Caki-casin' by Early Bilinguals.
Proceedings of the Annual Boston University Conference on Language Development 30,
1: 305-315.
Kwon, Nayoung, and Maria Polinsky. 2005. Heritage language retention: A quantitative
study. Heritage Learners: Overcoming Curricular Challenges. Paper presented at the
Symposium of the 25th meeting of the South Asian Language Analysis (SALA)
roundtable, September 16-18, 2005.
http://ling.ucsd.edu/~nayoung/Papers/SALA_slides_Heritage%20Language%20Retentio
n.pdf
Lee, Dong Jae. 2001. Recent trends in foreign language teaching in the United States:
The role of heritage learners. The Korean Language in America: Volume 6. Papers from
the Annual Conference and Teacher Training Workshop on the Teaching of Korean
Language, Culture, and Literature, ed. by Jungho Ree (6th, Manoa, Hawaii, August 2-5,
2001). 203-211.
Leisio, Larisa. 2001. Morphosyntactic convergence and integration in Finland Russian.
Ph. D. Diss., University of Tampere.
Levine, Glenn. 2000. Incomplete First-Language Acquisition in Bilingual Situations:
Yiddish in the United States. Linguistische Arbeiten 426. Max Niemeyer Verlag.
Miyagawa, Shigeru. 2006. Locality in syntax and floated numeral quantifiers in Japanese
and Korean. Proceedings of the 14 th Japanese/Korean Linguistic, ed. by Timothy J.
Vance and Kimberly Jones, 270-282. Stanford, CA: CSLI.
Montrul, Silvina. 2002. Incomplete Acquisition and Attrition of Spanish Tense/Aspect
Distinctions in Adult Bilinguals. Bilingualism: Language and Cognition 5. 39-68.
Montrul, Silvina. 2004a. The Acquisition of Spanish. Morphosyntactic Development in
Monolingual and Bilingual L1 Acquisition and in Adult L2 Acquisition. Amsterdam:
John Benjamins.
Montrul, Silvina. 2004b. Subject and Object Expression in Spanish Heritage Speakers: A
Case of Morphosyntactic Convergence. Bilingualism: Language and Cognition 7. 125-
142.
NVTC: National Virtual Translation Center: 2006. Language learning difficulty for
English speakers
43
http://www.nvtc.gov/lotw/months/november/learningExpectations.html.
Oh, Janet. S., Sun-Ah, Leah M. Knightly, and Au, Terry. K. 2003. Holding on to
Childhood Language Memory. Cognition 86.B53-B64.
Peyton, Joy K., Donald A. Ranard, and Scott McGinnis (eds.) 2001, Heritage Languages
in America: Preserving a National Resource. Washington, D.C., and Delta Systems,
McHenry, IL: Center for Applied Linguistics.
Polinsky, Maria. 1995. Cross-linguistic parallels in language loss. Southwestern Journal
of Linguistics 14. 87-125.
Polinsky, Maria. 1997. American Russian: Language loss meets language acquisition. In
Wayles Browne et al., eds. Formal Approaches to Slavic Linguistics, 370-407. Ann
Arbor, MI: Michigan Slavic Publications.
Polinsky, Maria. 2000. The Composite Linguistic Profile of Speakers of Russian in the
US. The Learning and Teaching of Slavic Languages and Cultures, ed. by Olga Kagan,
and Benjamin.Rifkin, 437-465. Bloomington, IN: Slavica.
Polinsky, Maria. 2003. Review of Zemskaja 2001. Heritage Language Journal 1.
http://www.heritagelanguages.org/
Polinsky, Maria. 2005. Word class distinctions in an incomplete grammar. Perspectives
on language and language development, ed. by Dorit Ravid and Hava Bat-Zeev
Shildrkrodt, 419-436. Dordrecht: Kluwer.
Polinsky, Maria. 2006. Incomplete Acquisition: American Russian. Journal of Slavic
Linguistics 14: 161-219.
Polinsky, Maria. in press a. Gender under incomplete acquisition. Heritage Language
Journal. http://www.heritagelanguages.org
Polinsky, Maria. in press b. Heritage language narratives. Heritage Language Education:
A New Field Emerging, ed. by Donna Brinton, Olga Kagan, and Susan Bauckus.
Mahwah, N.J: Lawrence Erlbaum Associates.
Polinsky, M.: in press c, Without aspect.
http://www.fas.harvard.edu/~lingdept/documents/Aspect_IA_001.pdf
Rickford, John R. 1987. Dimensions of a Creole continuum: History, texts & linguistic
analysis of Guyanese Creole. Stanford, CA: Stanford University Press.
Roca, Ana (ed.). 2000. Research on Spanish in the United States. Sommerville, MA:
Cascadilla Press.
44
Roosevelt, Theodore. 1918. The Square Deal in Americanism, in Theodore Roosevelt,
The Great Adventure. New York, NY: C. Scribner’s Sons.
Sanchez, Rosaura. 1983. Chicano discourse. Rowley, MA: Newbury House.
Seliger, Herbert W., and Robert M.Vago. 1991. First Language Attrition. Cambridge:
Cambridge University Press.
Silva-Corvalan, Carmen. 1994. Spanish in Los Angeles. Oxford: Clarendon Press.
Sohn, Ho-min. 2001. The Korean Language. Cambridge: Cambridge University Press.
Sorace, Antonella. 2004. Native language attrition and developmental instability at the
syntax-discourse interface: Data, interpretations and methods. Bilingualism: Language
and Cognition 7. 143-145.
Sundara, Megha, Linda Polka, and Shari Baum. 2006. Production of Coronal Stops by
Simultaneous Bilingual Adults. Bilingualism: Language and Cognition 9. 97-114.
Thal, Donna, Elizabeth Bates, Judith Goodman, and J. Jahn-Samilo. 1997. Continuity of
language abilities in late- and early-talking toddlers. Developmental Neuropsychology
13. 239-273.
Thal, Donna, Elizabeth Bates, Mary J. Zappia, and Melinda Oroz. 1996. Ties between
lexical and grammatical development: Evidence from early-talkers. Journal of Child
Language 23. 349-368.
Toribio, Almeida Jacqueline. 2001. Accessing Bilingual Code-Switching Competence.
International Journal of Bilingualism 5. 403-436.
Tsimpli, Ianthi, Antonella Sorace, Caroline Heycock, Francesca Filiaci, and Maria
Bouba. 2003. Subjects in L1 Attrition: Evidence from Greek and Italian Near-Native
Speakers of English. Proceedings of the Annual Boston University Conference on
Language Development 27. 787-797.
Tsimpli, Ianthi, Antonella Sorace, Caroline Heycock, and Francesca Filiaci. 2004. First
language attrition and syntactic subjects. International Journal of Bilingualism 8. 257-
278.
Valdés, Guadalupe. 1995. The Teaching of Minority Languages as 'Foreign' Languages:
Pedagogical and Theoretical Challenges. The Modern Language Journal 79. 299-328.
Valdés, Guadalupe. 2000. The teaching of heritage languages: An introduction for Slavic-
teaching professionals, ed. by Olga Kagan and Benjamin Rifkin. The Learning and
Teaching of Slavic Languages and Cultures, 375-403. Bloomington, IN.: Slavica.
45
Valdés, Guadalupe. 2001. Heritage language students: Profiles and possibilities. Heritage
Languages in America: Preserving a National Resource, ed. by Joy K. Peyton, Donald .A.
Ranard, and Scott McGinnis, 37-77. Washington, D.C., and Delta Systems, McHenry, IL:
Center for Applied Linguistics,
Valdés, Guadalupe, Sonia V. González, Dania López García and Patricio Márquez.
in press. Heritage Languages and Ideologies of Language: Unexamined Challenges.
Heritage Language Education: A New Field Emerging ed. by Donna Brinton, Olga
Kagan, and Susan Bauckus. Mahwah, N.J: Lawrence Erlbaum Associates.
Van Deusen-Scholl, Nelleke. 2003. Toward a definition of heritage language:
Sociopolitical and pedagogical considerations. Journal of Language, Identity, and
Education 2.3. 211-30.
Wang, Xun (ed.) 1996. A View from Within: A Case Study of Chinese Heritage
Community Language Schools in the United States. Washington, DC: The National
Foreign Language Center.
Weger-Guntharp, Heather. 2006. Voices from the Margin: Developing a Profile of
Chinese Heritage Language Learners in the FL Classroom. Heritage Language Journal 4.
http://www.heritagelanguages.org/
Wiley, Terrence G. in press. Chinese “Dialect” Speakers as Heritage Language Learners:
A Case Study. Heritage Language Education: A New Field Emerging, ed. by Donna
Brinton, Olga Kagan, and Susan Bauckus. Mahwah, N.J.: Lawrence Erlbaum Associates.
Wolfram, Walt, and Natalie Schilling-Estes. 1998. American English: dialects and
variation. Malden, MA: Blackwell.
Zapata, Gabriela, Liliana Sanchez, and Almeida Jacqueline Toribio. 2005. Contact and
Contracting Spanish. International Journal of Bilingualism 9. 377-395.
Zemskaja, E. A. (ed.) 2001. Jazyk russkogo zarubežja: Obščie processy i
rečevye portrety [Language of the Russian diaspora: Common processes and speech
portraits].Moscow/Vienna: Jazyki slavjanskoj kul’tury/Wiener slawistischer Almanach.
Xiao, Yun. 2006. Heritage learners in the Chinese language classroom: Home
background. Heritage Language Journal 4.
http://www.heritagelanguages.org/
