Language Choice on Soc.Culture.Punjab
******* PAOLILLO ******** EJC/REC Vol. 6, No. 3, 1996 ******
LANGUAGE CHOICE ON SOC.CULTURE.PUNJAB
John C. Paolillo
University of Texas at Arlington
Abstract. This article examines factors
influencing language choice on the Usenet
newsgroup soc.culture.punjab, a forum chartered
for discussion of the culture of the Punjab region
of India and Pakistan. English and Punjabi are
both used on soc.culture.punjab, although English
is statistically predominant, and the majority of
Punjabi words are isolated borrowings occurring in
otherwise English messages. The results of a
functional analysis of code-switching reveal that
Punjabi is used for marked discourse functions
(often negatively), and only with interlocutors
who are Punjabi community members. This limited
usage is accounted for in terms of four factors
which disfavor Punjabi use: inter-generational
language shift, cultural ambivalence among
expatriates, the prestige status of English in
South Asia, and the predominance of English on the
As computer networks expand to link ever more remote
places, we witness the evolution of a language contact
situation of an unprecedented scale. The Internet now links
millions of people with hundreds of different native
languages worldwide, and these users bring with them highly
varying language backgrounds. At times this linguistic
diversity is manifest in a user's limited proficiency in a
major world language such as English. Other times it may be
manifest in the use of the user's native language on-line,
especially in "virtual communities" (Jones, 1995b;
Rheingold, 1993) made up of members of the user's native
When different language groups come into contact, a
protocol must be negotiated, via collective consensus or
inter-group conflict, which establishes the ways that the
different languages will be used. These patterns of
language use both reflect and influence the status of the
contributing language groups (Eastman, 1992). In a contact
situation, a number of outcomes are possible. For example,
a dominant language may completely replace other languages.
Alternatively, a dominant language may be used for some (or
most) communicative functions, while one or more non-
dominant languages may be used for others. Finally, both
dominant and non-dominant languages may be used together, in
sequential alternation, in what is known as "codeswitching"
The Internet offers an excellent opportunity to observe
the processes by which the relative status of different
languages are negotiated in contact situations, both because
extensive contact between groups is taking place, making
negotiation necessary, and because the discourse by which
language status is negotiated can be readily collected,
archived and analyzed. How frequently are languages other
than English used on-line? What patterns of use -- e.g.,
codeswitching and/or functional specialization of languages
-- are evident in the on-line discourse of bilinguals, and
what factors motivate these patterns?
In this study, I address these questions for one
community of users, Punjabi expatriates living in Canada,
the UK and the US. This community is interesting for
several reasons. First, South Asia is well represented on
the Internet, especially on Usenet newsgroups and IRC
(Internet Relay Chat) channels. In July of 1996, 23 out of
145 total newsgroups in the soc.culture hierarchy were
devoted to discussion of South Asian cultures. Regularly
maintained South Asian IRC channels include #india, #punjab,
#pakistan, #bengali, #srilanka, #tamil, and many others. Up
to the present time, almost all of the participants in these
fora have been expatriate South Asians, even though Internet
access is available in South Asia through such networks as
ERNET in India. However, many expatriate users are
relatively recent immigrants and have a South Asian language
-- in this case Punjabi -- as their native language.
Therefore we might expect that considerable use of Punjabi
would be found on-line, expecially where expatriates gather
in virtual communities of their peers.
At the same time, the sociolinguistic situation of
expatriate Punjabis, like that of other South Asians, is
complicated by contact with English. Since the expatriates
are often affiliated with colleges and universities where
English is the language of instruction, they can be assumed
to have considerable exposure to and proficiency in English.
However, their background in Punjabi language and education
varies greatly, from those who have received higher
education in Punjabi, to those who emigrated as young
children to English-speaking countries and may have only
passive proficiency in Punjabi. Nor does the prominence of
English begin with members' arrival in the expatriate
setting. In India, English is the language of a former
colonial administration, and knowledge of English and
English education continue to have special status (Fasold,
1984). English is favored over regional languages for
inter-ethnic communication, since it is regarded as
ethnically neutral (Fasold, 1984; Pandharipande, 1992).
Moreover, economic and social advancement often go hand in
hand with and/or result from greater proficiency in English,
and members of the expatriate community tend to come from
the more advantaged sectors of South Asian society. Thus it
is likely that the participants on soc.culture.punjab all
had extensive exposure to English even before immigrating to
countries where English is the national language. These
factors might lead us to expect, in conflict with the
expectation stated above, that Punjabis would use a high
proportion of English, and possibly only English, in their
on-line communication. An additional goal of this research,
therefore, is to determine how these conflicting situational
factors translate into actual language choice in a computer-
Data and Methodology
The on-line Punjabi expatriate community first
expressed itself in the formation of a Usenet newsgroup in
September 1994. Usenet is an asynchronous mode of computer-
mediated communication in which postings similar to e-mail
messages are temporarily archived and forwarded from site to
site. Messages are generally public, and are read using a
"newsreader" program, which typically organizes the messages
by subject header line, or "thread", and by time and date of
receipt within each thread. Normally, anyone can post a
message, by using the same software used to read messages.
Some moderated groups exist on Usenet, in which the messages
are forwarded to a moderator who selects which ones are
posted, but soc.culture.punjab and most other South Asian
newsgroups are unmoderated. Changes in status of Usenet
newsgroups, e.g., the creation of new groups, the moderation
of unmoderated groups, etc., are generally accomplished
through popular vote by e-mail.
The initial participants on soc.culture.punjab were
mostly drawn from soc.culture.indian, a Usenet group with a
broad charter for the discussion of Indian culture, but
which is largely dominated by discussions of Indian national
and South Asian regional politics. At the time of the
formation of soc.culture.punjab, the population of South
Asians reading Usenet had been undergoing rapid growth,
following an overall trend in Usenet at the time. This
growth brought about a re-organization of South Asian Usenet
groups, in particular the formation of several other
newsgroups with regionally-specific South Asian foci, such
as soc.culture.kerala, soc.culture.tamil, etc. At the time
of my initial observations, soc.culture.punjab had just been
created through popular vote for the discussion of the
language, literature, arts, music, politics and culture of
the Punjab region of India and Pakistan.
I began my observation by systematically downloading
all messages posted to soc.culture.punjab on a daily basis
during two separate periods, September 20 to October 13,
1994, and February 6 to 23, 1995. I then stripped header
information and extended quotations of previous messages
from each of the messages. The resulting data set consisted
of 759 messages, for a total of 81,000 words. Since Usenet
posters cannot tell who is reading Usenet postings (other
than if someone responds by posting or e-mail), my
observation of the group did not influence the posters'
In order to analyze patterns of language choice in the
data, I constructed a concordance for the entire corpus
using Conc, a Macintosh program, which was used to produce
word frequency lists and summary statistics. The Punjabi
portions of the corpus, whether individual words or longer
stretches, were isolated and collected in a separate corpus
so that frequency statistics could be prepared for Punjabi
alone. I also categorized all instances of non-English
expressions, including loanwords (i.e., words borrowed from
other languages). Longer uses of more than one word were
categorized according to their creativity (whether the
poster was producing unique Punjabi utterances, or re-using
formulaic or widely quoted utterances), and their
communicative function. Following this, I compared the uses
of Punjabi with the most common uses of English on the
newsgroup. Finally, I interpreted my findings in light of
meta-discourse about language choice on soc.culture. punjab
as represented in the corpus and in messages posted from
March to May 1995.
Statistical Distribution of English and Punjabi
The first question I addressed was: How often do
participants on soc.culture.punjab use Punjabi? Table 1
presents the rates of English, Punjabi, and mixed English
and Punjabi use by message.
Table 1. Use of Punjabi and English, by message
Only English 84 (11.1%)
Some English, Some Punjabi 655 (86.3%)
Only Punjabi 20 ( 2.6%)
Use of only English is more than four times as great as use
of only Punjabi. However, the overwhelming majority of
posts (86.3%) use both English and Punjabi. Since the only
English and only Punjabi posts make up less than 14% of the
corpus, it might seem that the use of English and Punjabi is
more or less balanced. This appearance evaporates when the
relative frequencies of Punjabi and English words in the
corpus are considered.
Table 2. Use of Punjabi and English, by words
Tokens Types Ratio
Punjabi 11900 (15%) 4560 (33%) 0.38
English 69350 (85%) 9340 (67%) 0.13
Total 81250 13900 0.17
English words (tokens) are more frequent than Punjabi words
in the corpus by a ratio of more than 5:1. Moreover, the
type/token ratio for Punjabi is almost three times that for
English. The reason for this is that the overwhelming
majority of Punjabi tokens appear as loanwords in English
sentences using English function words, and not in sentences
framed with Punjabi grammar using Punjabi function words.
The most frequent tokens in any corpus of connected text are
the grammatical or "function" words of a language
(Tesitelova, 1992). Since several of the same group of 30
or so function words may appear in any given sentence,
function words have very high frequencies and they bring
down considerably the overall type-token ratio for a corpus.
This interpretation is supported by the information in
Table 3. Among the 30 most frequent types in the complete
corpus (comprising 34,383 tokens, or 42% of the tokens in
the corpus) we find only two Punjabi words: _punjab_ (rank
26, 456 tokens), and _punjabi_ (rank 28, 409 tokens). The
remaining 28 forms are all English function words.
Table 3. 30 most frequent types for the complete corpus
rank freq type rank freq type rank freq type
1 5033 the 11 1013 you 21 563 from
2 2923 of 12 843 for 22 528 with
3 2663 to 13 832 are 23 478 all
4 2410 and 14 830 this 24 462 or
5 1973 a 15 710 not 25 460 by
6 1948 in 16 701 have 26 456 punjab
7 1765 I 17 654 be 27 444 they
8 1535 is 18 643 was 28 409 punjabi
9 1301 that 19 621 on 29 408 but
10 1019 it 20 602 as 30 396 my
The third most common Punjabi word (rank 33, 384 tokens) is
_Singh_, a Punjabi surname. All three are easily classified
as loanwords on the independent grounds that their use in
English is necessitated by talking about the Punjab region,
Punjabi, and Punjabi people. Note that no Punjabi function
words appear among the highest ranking forms. The
statistical profile of the corpus is that of an English
discussion in which Punjab and Punjabi people figure heavily
Even when we consider the Punjabi corpus separately, we
find that function words do not predominate in the way that
the English function words do. Nearly half (47%) of the
Punjabi corpus is accounted for by names or components of
names (e.g., the city names _Lahore_ and _Amritsar_).
Moreover, the Punjabi function word types do not cluster
together among the top frequency range of the Punjabi
corpus, as the corresponding English forms do in the
complete corpus. Instead the 37 function word types are
sprinkled throughout the top 100 ranks, while 28 function
word forms occur in the top 30 ranks for the complete
corpus. The 30 most frequent types in the Punjabi corpus
are listed in Table 4; in the glosses, square brackets 
indicate function-word items, while personal names are
indicated by .
Table 4. 30 most frequent types for the Punjabi corpus
rank fr. type gloss rank fr. type gloss
1 456 punjab Punjab 16 32 ji [honorific]
2 409 punjabi Punjabi 17 32 suminderpal
3 248 singh 18 31 sat truth
4 79 guru teacher 19 30 ke [that]
5 68 jatinder 20 30 sandeep
6 52 rajbir 21 30 ve [also]
7 47 harpreet 22 28 kalia
8 44 te [and] 23 27 akal God
9 39 nanak 24 26 di [genitive]
10 36 hai [is] 25 26 lahore Lahore
11 35 de [genitive] 26 26 ne [by]
12 34 deol 27 25 na [not]
13 34 khalsa 28 24 amritsar Amritsar
14 33 bajwa 29 24 bhangra type of music
15 33 da [genitive] 30 24 kaur
Table 4 shows that only 11 of the 30 most frequent Punjabi
types are function words (as compared with 28 English
function words out of the 30 most frequent words for the
entire corpus), and 12 are names (as compared with 0/30).
These findings are consistent with the claim that a large
number of Punjabi words appear in English sentences as
borrowings rather than as parts of all-Punjabi sentences.
There is thus a clear asymmetry in the extent to which
English and Punjabi are employed on soc.culture.punjab, with
English being used more often and more productively. Why
should this asymmetry exist? Alternatively, why should the
Punjabi language be used on soc.culture.punjab at all, given
that it is used productively so little? To answer these
questions, we turn to a qualitative analysis of the
functions of Punjabi in soc.culture.punjab discourse.
Functions of Punjabi
Two principal categories of uses of Punjabi emerged as
the result of my analysis of postings on soc.culture.punjab:
"fixed" and "creative" uses. Fixed uses involve the use of
Punjabi sentences and phrases which are essentially
invariant from occasion to occasion. Creative uses involve
the productive application of the linguistic resources of
Punjabi to produce novel, situationally relevant and
appropriate utterances. These two types of uses can in turn
be further sub-categorized, according to the communicative
function of the utterance. Below, I first describe the sub-
types of fixed uses, followed by the sub-types of creative
uses. These uses are then compared with the most common
uses of English on the same newsgroup.
Punjabi is used in a large number of functions in which
the expressions are essentially invariant. There are two
main types of fixed expressions: (i) those which are
formulaic, and (ii) those which are rehearsed or quoted.
Formulas. Fixed formulas commonly appear in the
openings of messages, and also in the signature files
appearing at the ends of many messages. Some of these are
highly conventional, as in the traditional Sikh greetings in
(1) and (2).
1) Sat Sri Akal
"God is truth" (= "hello")
2) Waheguru Jee Ka Khalsa
Waheguru Jee Ki Fateh!
"The Khalsa (Sikh) belongs to God,
Victory belongs to God."
Others may bear a mark of either the religious or political
affiliation of the author of the message, as in examples (3)
3) KHALISTAN ZINDABAD
"Long live Khalistan!"
(Khalistan = homeland sought by Sikh separatists)
4) RAJ KAREGA KHALSA
"The Khalsa (Sikhs) will govern."
Occasionally a new formula is created from old ones through
a process of minimal substitution, as in the sarcastic reply
to (3) found in (5).
5) KHALISTAN MURDHABAD
"Long may Khalistan remain dead!"
In any case these formulas are only created once; they are
recycled in toto for subsequent uses, either as memorized
phrases, or even more mechanically by being incorporated
into a poster's signature (.sig) file which is appended
automatically to each posting.
Rehearsed. The second type of fixed expression is
typically associated with verbal art, especially poetry. In
some cases, a poem or song appears to be reproduced from
memory, since it is only partially complete. Other times it
appears to have been typed in using a book or other
reference as a guide. The example below illustrates both
possibilities. (English translation of the Punjabi in the
examples below has been added within square brackets.)
Subject: Re: Nanak Naam Jahaz
Date: 27 Sep 1994 19:01:59 GMT
: " De Ey Shiva bar moh hae eha
: Subh Karman se kab ho na daro
: Na Daro...... "
: Can someone please complete and post this
: exceptionally inpirational
deh Shiva bar mohe ehai
shubh karman te kab_hoon na Tron
na tron ar so jab jae laron
nishchey kar apni jeet karon
ar sikh_ho apney hi man kau
eh lalach hau guN tau uchron
jab aav ki audh nidhan banai
at hi raN mai tab joojh maron
[ Grant unto me this boon, O Lord, ]
[ That I may never be deterred from doing good deeds. ]
[ I should have no fear of the enemy when I go to battle,]
[ And turn victory decidedly to my side. ]
[ In my mind there is but one desire ]
[ That I may ever be singing Thy praises. ]
[ And, when the time comes, I should die ]
[ Fighting in the thick of action.  ]
Excuse the typing mistakes, if any.
When song lyrics or poetic literature are quoted in
this fashion, they typically have at least some words
glossed. It is also common for people to ask for a poem to
be completed (as in the above example), or to ask for the
meaning of a word, stanza or the poem itself, suggesting
that many members of the group do not fully control the
language of the poems.
What is noteworthy about the two types of fixed, non-
creative Punjabi use is their frequency: formulaic Punjabi
expressions occur in 12% of the posts overall, and many
posters ritually open or close with a formulaic expression.
Punjabi poetry is quoted in 8% of the posts overall, and
posting poetry is the major form of Punjabi message on the
newsgroup: posts containing poetry have 52% Punjabi tokens
on average, and 50% of the messages with 80% or more Punjabi
tokens contain poetry. Thus we find that both the most
frequent uses (formulaic salutations) and the most sustained
uses (quoting poems and lyrics) of Punjabi involve fixed,
invariant Punjabi expressions.
Punjabi is also used creatively on soc.culture.punjab,
though less commonly than in its fixed uses. Three creative
uses are identified below: insults, appeals, and the climax
of a joke.
Insults. Most prominent among the creative uses of
Punjabi is its use in confrontations, especially in insults.
Usenet interactions are often confrontational, and a variety
of strategies for verbally challenging others may be
observed in different newsgroups (MacKinnon, 1995;
McLaughlin, Osborne & Smith, 1995). On soc.culture.punjab,
if a conflict escalates sufficiently, Punjabi may be used as
a means of strengthening a confrontation or challenge. The
message below contains an example of a challenge embedded in
the quoted message, partly in Punjabi, responded to also
Subject: Re: FOR SUMINDER AND RAJWI ONLY
Date: Thu, 16 Feb 1995 21:55:31 EST
In article <email@example.com>,
firstname.lastname@example.org (Harpreet Singh) writes:
>OYA VIR LOOOOG [Hey, brothers! ]
thanks for saying veer [brother]
>TUCI PAGAL HO KAYA... [You have gone insane... ]
>SUMINDER TO THE BOLA CI KEY POST [Suminder you also have ]
[said you (won't make) ]
[any posts... ]
too or tu [you (alt. spellings) ]
tee [also ]
booleaa [said ]
kee [that ]
>NAHI KAYEJA..THE HUM KE HUYA [won't make.. ]
[What has happened now? ]
kareen gaa [will make]
>SUMINDER U ARE NOT A MAN OF UR
>OWN WORDS....I THINK U HAVE A LOOSE
>SCREW IN UR HEAD...TELL U WHAT ..WHY
your are right there
>DON'T U ASK YOUR MOTHER
>IF..YOUR HEAD OR
haayee besharam [Oh, shameful. ]
too koee apnee mummy too [You are asking this kind]
edda dee sawaal poochdaa hai [of question of anyone's ]
>YOUR FEET CAME UP FIRST.
>I DON'T THINK U ARE
>A NORMAL BORN PERSON..
>AND VIR RAJWI TU THE PUDDDU HAI [And brother Rajwir, you ]
[are idiotic. ]
>PUDDDU HE RAHGA [constantly being idiotic]
raheengaa [(always) being]
harpreet kidda daa punjabi hai [This is Harpreet's kind ]
too punjabi vee naee [of Punjabi. You don't ]
boolnee aaundee [even know how to speak ]
teraa veer [your brother]
The post above occurs at the climax of a heated
exchange between two posters, Harpreet and Suminderpal,
following which Suminderpal refrains from posting for a
time. Here, Suminderpal is responding to Harpreet, who has
escalated the exchange by resorting to Punjabi to deliver
his most effective insults. Not to be outdone, Suminderpal
responds point-for-point, starting in English and then
switching into Punjabi. Some of his responses to Harpreet
are "corrections" of his spelling, a common tactic in Usenet
flame wars in English (MacKinnon, 1995). Here, the tactic
has the effect of disarming Harpreet's use of Punjabi.
However, Punjabi is normally written in a non-roman script
that can't be reproduced easily using ASCII characters. The
irony of Suminderpal's response is that one can't really
"correct" romanized Punjabi spelling; there are several
different romanization schemes for Punjabi, and Harpreet's
spelling merely represents one.
Appeals. The role of Punjabi in the confrontations
need not be directly confrontational. In the post below,
also in response to the thread that gave rise to the above
exchange, the poster takes sides and identifies with one of
the feuding parties, through an appeal framed in Punjabi.
From: email@example.com (SEONG KANG)
Subject: Re: SUMINDERPAL SINGH BEDI IS A ...
Date: Wed, 15 Feb 1995 18:03:03 GMT
In article firstname.lastname@example.org
(Harpreet Singh) writes:
>MAN OF WHIS WORDS..
>HE SAID HE WILL NOT COME BACK AND HE IS TILL NOT BACK
>THATS GREAT, KEEP IT UP
Oh Brha wapas ah! Bhapa ji Choka ka kaha nahi munne da
[Oh, brother, come back! Don't pay attention to what he
Climax of a joke. Another function of Punjabi is found
in jokes. In the post below, the butt of the joke (Sardar
Bhanta Singh) is caricatured as ignorant by his use of
accented English with codeswitching into Punjabi. The
character's utterance is the punch-line, and thus the
climactic peak of the joke.
From: email@example.com (Top Gun)
Subject: Re: The Phantom Strikes Again
Date: 12 Oct 1994 07:20:42 GMT
I am back! Thanks for your support. I abandoned my
"sardar of waterloo" thread since a few people did not seem
to appreciate my jokes. This is only for those with a sense
NOTE : The following is a true story. All names have been
changed to protect the innocent....
Ok, so there was this archaeology conference in geneva.
Experts came from all over the world. And naturally, the
hero of our story, Sardar Banta Singh of Whaaterloo was
among the great names in science to be invited. During
dinner, he got to sit next to two famous archaelogists, one
russian and the other American. While our hero was busy
devouring the tandoori chicken, the other two started to
brag. The Russian says "while digging thru an ancient ruin
in Moscow, i came across some thick cables, which showed
that the russians had the telegraph system long b4 it was
invented." Banta Singh looked unimpressed and picked another
piece of chicken. The American, not to be outdone, says,
"while digging thru a ruin in New Mexico, I found thin
cables which clearly show that we had telephones long b4
they were invented." Now Sardar Banta Singh puts down his
chicken and decides to speak. "Jadon mein hoshiarpur wich
upni digging kar reha sigaa, I phound no cablaan tey no
wiraan. [When I was doing my own digging in Hoshiarpur, I
found no cables and no wires.] Thees is vairy indicative
aaf the phact that the sardars [a Punjabi caste] had the
moost sofisticated wireless system!"
-Berzerk in Berkeley
Such codeswitching would be appropriate in a Punjabi ethnic
conversational context, but is inappropriate in the high-
status international context of the situation represented in
the joke, especially given that the American and Russian
addressees are unlikely to understand Punjabi. Thus the use
of Punjabi in the joke is associated with inappropriateness
and other socially undesirable qualities.
Some general observations can be made about the
creative uses of Punjabi in the corpus. First, Punjabi is
never creatively used to express a neutral affect. Whether
positive or negative, its use is always "marked" in some way
to express special meaning, whether it be emphasis or in-
group membership. This type of code choice pattern is
characteristic of situations in which expatriate speakers of
one language are shifting to another (Myers-Scotton, 1993b,
Second, Panjabi is only creatively used with in-group
members. There is a small core group of posters who
distinguish themselves through more frequent Punjabi use,
and who mutually exchange Punjabi creatively. Other posters
not in this core group, even if they are regular posters
otherwise knowledgeable about Punjab and Punjabi culture,
are seldom participants in such exchanges. According to the
posters' own comments, recency of residence in the Punjab
region of India and Pakistan appears to play a role in this
"core" group membership.
Third, posters may be ridiculed for lack of Punjabi
knowledge, either for displaying poor control of it (e.g.,
through spelling), or for code-mixing Punjabi and English.
As a consequence, using Punjabi creatively in a posting
constitutes a risk. An example of this type of situation is
the first example of creative Punjabi use given above, in
which Suminder corrects Harpreet's Punjabi. The risk of
criticism may deter posters from using Punjabi, especially
if they speak a non-standard variety, or if they are second-
generation speakers whose knowledge of the language is weak.
Functions of English
In order to understand fully the marked nature of
Punjabi use on soc.culture.punjab, we need to ask what the
nature of the "unmarked" discourse is on soc.culture.punjab.
That is, what are the functions of English? In this section
I identify two predominant uses: political argument and
Political argument. The most salient form of discourse
on soc.culture.punjab can be broadly described as "political
argument". Views expressed in these posts reflect a variety
of different political, social and religious perspectives on
Punjabi society. The most common recurrent subject of these
postings is the partition of the Punjab region, and the
political federation of East and West Punjab with India and
Pakistan respectively, as illustrated in the post below. In
addition to the partition/federation issue, there are
religious divisions (Sikh, Muslim, Hindu), social caste
divisions (Brar, Jat, Gill, etc.), and political ideologies
(democratic, socialist, theocratic) which intersect in
Subject: Re: Khalistan Re: weltpolitik
Date: 10 Oct 1994 02:08:33 GMT
> In case of Punjab, the demand for autonomy should
> and must be at the forefront of Punjabi political
> agenda.By realizing the inherent cracks within this
> Indian mosaic, we must recoginize the need to
> delink our political and economic relationship with
> the Central government.
I disagree. An independent Punjab would be an non-viable
state. How could it exist land-locked between 2 rival giants
- India & Pakistan? (Please don't tell me that Khalistan
includes the Gujarati coastline) And how do your arguments
apply to Pakistan. After all, Pakistan is also a multi-
ethnic state whose largest province is the larger portion of
Strikingly, political argument on soc.culture.punjab is
conducted entirely in English. One reason for this is that
some of the political postings are cross-posted to other
South Asian newsgroups (typical groups are
soc.culture.indian, soc.culture.pakistan and
soc.culture.afghanistan) and therefore have an inter-ethnic
audience. English, being the language of the former
colonial administration over most of South Asia, has the
status of an inter-ethnic lingua franca (Fasold, 1984;
Pandharipande, 1992). This situation automatically selects
for English as the common code where an inter-ethnic
audience is likely, e.g., in cross-posts. Also, because of
its association with education and government, i English has
greater prestige than any vernacular. Thus, both because of
their inter-ethnic audience and their association with
government, political arguments overwhelmingly occur in
Insults. Viewpoint differences evident from ongoing
political arguments often form the basis for taunts and
insults. The Punjabi insults illustrated earlier
notwithstanding, English is the preferred medium for
insulting on soc.culture.punjab, unless a poster's Punjabi-
ness has been called into question, in which case both
Punjabi and English insults may be used. Flame wars on
soc.culture.punjab always begin in English, and only
sometimes escalate into Punjabi. Once flaming commences in
Punjabi, it often does not last more than a couple of
additional posts; there does not seem to be any further
stage of escalation once Punjabi has been used.
Other uses of English. There are, of course, other
creative uses of English on soc.culture.punjab, for example
in discussing recent films or sporting events. We also find
fixed uses such as formulaic phrases and quoted English song
lyrics in message signatures, much as we do for Punjabi. In
fact, any communicative function other than symbolizing
Punjabi identity may employ English. Thus English is the
default language for all communicative purposes on
Summary and Interpretations
The conditions on the creative use of Punjabi on
soc.culture.punjab relegate the Punjabi language to a
relatively minor role. Punjabi is used to serve the group-
identification needs of the Punjabi community on-line, but
this is adequately accomplished most of the time with
minimal, formulaic uses of Punjabi, and Punjabi words
borrowed into English. Creative uses of Punjabi also serve
group-identification purposes, but only for a minority of
posters on soc.culture.punjab, who arguably achieve a
special status on the group through this usage. In
addition, creative use of Punjabi carries higher and
potentially unacceptable risks for less fluent posters.
The principal pattern of code choice on soc.culture.
punjab is one in which English is "unmarked" -- it is used
for the most general range of functions -- and Punjabi
"marked" -- it is used for special, emphatic communicative
effects. In Myers-Scotton's (1993a, 1993b) terms, English
functions as the "matrix language" for code choice/
switching, both for the newsgroup as a whole and in the
overwhelming majority of individual interchanges.
The situation on soc.culture.punjab, however,
represents an exaggeration of these tendencies, to the point
where Punjabi is functionally marginalized and apparently
avoided altogether by some. By "functionally marginalized"
I mean that the language is used by members of the community
in fewer and fewer communicative functions (cf. Hill,
1993). In what follows, I suggest that the functional
marginalization of Punjabi on-line is a consequence of four
interrelated factors: inter-generational language shift,
cultural ambivalence of expatriate Punjabis, the prestige
status of English in South Asia, and Usenet language norms
that favor English. Each of these factors is discussed in
Inter-generational Language Shift
Second-generation children of immigrants in communities
undergoing language shift avoid the mother tongue in favor
of the second language, and as a consequence may fail to
acquire the mother tongue or may acquire it only partially.
In such a situation, a distinction between creative and
fixed uses is important, since only creative uses constitute
evidence that active knowledge of the language is being
passed down from the older generation to the younger.
Many posters on soc.culture.punjab are second-
generation Punjabis, as indicated in the content of their
messages to the group. The limited extent of creative usage
of Punjabi, together with many posters' self-reported
insecurity in using Punjabi, suggest that active knowledge
of the language is not being passed down. Even some of the
fixed uses illustrate a lack of Punjabi knowledge among
certain posters. For example, one person posted a sardonic
variation on a common formula, triggering a flood of other
postings. Some posters wanted to ascertain if the original
poster's expression was correct or in jest, others asked
what the true expression was supposed to be, and others
offered opinions of what it should have said. This lack of
knowledge can be taken as evidence of language shift.
Further evidence of language shift is the fact that the
theme of inter-generational differences, including
differences in language use, is prominent in the topics
discussed on the newsgroup. An overwhelming number of
threads on soc.culture.punjab are concerned with issues of
cultural maintenance faced by Punjabi expatriate communities
in the US, UK and Canada, where most of the posters reside.
Topics range from interracial marriage to the practice of
Sikhism in the West, to worldwide Punjabi migrations, to
learning and using Punjabi. The picture that emerges from
these discussions is one in which expatriate Punjabi
communities are struggling to come to grips with a tendency
towards cultural (including language) shift, as second
generation Punjabis assimilate to the English-speaking
cultures in which they were born.
Limited proficiency in Punjabi has a clear impact on
code choice for many second generation participants on
soc.culture.punjab. However, this factor alone is not
sufficient to explain the choice of English as the default
language on the newsgroup. Other participants know Punjabi
quite well, e.g., because they are first generation
speakers, and yet they still use more English than Punjabi.
To understand this, it is necessary to consider the cultural
values and attitudes that posters associate with their
choice of codes.
Cultural Ambivalence of Expatriate Punjabis
Previous ethnographic research on expatriate Punjabi
Sikhs in California's Central Valley (Gibson, 1988)
characterizes them as having a strategy of cultural
accommodation without assimilation, i.e., as adapting to the
expatriate cultural context while maintaining a separate
Punjabi identity. However, although the Central Valley
Punjabi community as a whole places a value on cultural
maintenance, the attitudes of those in the second generation
toward American mainstream culture lean more toward
assimilation than those of their immigrant parents, as noted
above for soc.culture.punjab. This tension between cultural
maintenance and shift creates an ambivalence among
expatriates that affects both first- and second-generation
Punjabis, and has implications for language choice.
Two threads on soc.culture.punjab occurred during my
periods of observation that directly attest to cultural
ambivalence among members of the on-line Punjabi community.
The first thread, entitled "Punjabis who speak Hindi are
sellouts!", concerns the status of Punjabi vis-a-vis Hindi
and Urdu, the dominant languages spoken in India and
Pakistan, respectively, both related to Punjabi in structure
and origin. The thread was initiated by the following post,
in which the poster opines about a putative language shift
taking place in Haryana in India and in Punjab in Pakistan,
toward Hindi and Urdu, respectively.
From: firstname.lastname@example.org (Bajaj)
Subject: Punjabis who speak Hindi are sellouts!
Date: 11 Mar 1995 15:18:09 -0500
To any Punjabi out there, whether you are a Hindu, Muslim or
Sikh: If you speak Hindi(Urdu) instead of Punjabi, you are a
serious disgrace to your culture and you shouldn't call
yourself a Punjabi! If you speak Hindi or Urdu instead of
Punjabi, you sound like a wussy. If you think you sound
sophisticated, you are seriously having delusions!
This poster strongly connects Punjabi cultural identity
with Punjabi language maintenance, such that true
Punjabiness precludes use of Hindi or Urdu. Ironically, the
post is in English, a fact which was commented upon by many
other posters, who chided the original poster for his double
standard. He responds to this charge in a subsequent post
by reiterating his earlier position and commenting on the
use of English:
I have not mentioned anything against the use of
English. And on the internet English would be the
best way to get my message across.
Even though the poster is in the company of Punjabis,
he is communicating across ethnic lines, according to his
own post, to speakers of Hindi or Urdu, and so would need to
accomodate to these speakers in order to be understood. His
attitude toward English appears to be one of positive
In contrast with this positive pragmatism, other
Punjabis feel that the use of English with Punjabi
interlocutors conflicts with the maintenance of Punjabi
cultural identity. This is illustrated by the following
Subject: Re: Punjabis who speak Hindi are sellouts!
Date: 16 Mar 1995 22:31:24
Actually, the sellouts are the ones who think Punjabi is not
good enough for them and so they speak in English...even
when it's inconvenient.
Here's an example. I was talking with someone in Punjabi
when a desi [Punjabi ethnic] girl walked by. All of a
sudden, he switched from Punjabi to English. Or...try
talking to a Punjabi who has been in the States for a while,
and they'll try to show their "high" status by not talking
in Punjabi with you. Or...ask the proud parents who
proclaim their children only speak English.
According to this view, people who attempt to use English to
obtain benefits within the Punjabi community undermine their
Punjabi identity; that is, they are "sellouts".
The second thread, entitled "English in Gurudwaras",
concerns the use of English in Sikh temples for religious
rituals and observances. The majority of Indian Punjabis
are followers of the Sikh religion, the sacred scriptures of
which are written in an older form of Punjabi. A number of
expatriate communities have established Sikh temples, or
Gurudwaras, for the practice of their faith, where the
rituals are usually conducted in Punjabi. The "English in
Gurudwaras" thread was initiated by a poster who felt that
he would understand Sikhism, and hence his own cultural
identity, better if Sikh religious services were conducted
From: Kirn Braich
Subject: Re: Punjabi in Gurudwaras
Have you ever noticed that the average age in gurudwaras is
increasing all the time? Young people simply aren't going
to gurudwara because they can't understand the language.
If you and others hold on to your sanctimonious dream of
Punjabi in the gurudwaras at all costs, then you will
preside over the demise of Sikhism in the West. In your
model, the only attendees at gurudwara will be recent
immigrants from India and elders.
Be realistic. We live in the US, UK, and Canada, where the
children learn primarily English. Yes, you are correct that
the parents should teach children Punjabi. But the fact is
that most children raised here don't learn it fluently
enough to understand what goes on in gurudwaras. If you
*realistically* want to get young people (raised here) into
gurudwaras, then stop putting impediments in their way.
Anyway, Sikhism is about religious philosophy, not about
a certain language. Right?
This point of view was not universally accepted, however, as
many Punjabi Sikhs regard the Punjabi language to be
inseparable from the transmission Sikhism, and resent the
encroachment of English there. Thus, two central Punjabi
Sikh values, language and religion, were pitted against one
another in this thread.
The principle of cultural ambivalence helps to explain
why English and Punjabi are both used on soc.culture.punjab.
English is used as the default code that all posters will
understand. However, using English alone does little toward
expressing one's identity as a Punjabi. Thus, posters find
it desirable to use Punjabi occasionally, in order to
express their Punjabi identities.
For non-fluent posters, however, posting in Punjabi
risks undermining the Punjabi identity they want to express.
In using Punjabi, they must face the twin undesirable
possibilities of being ridiculed for insufficient knowledge
of Punjabi, or (if their use happens to be more fluent) of
being associated with values that they do not hold. Yet
failure to use Punjabi would not properly acknowledge their
Punjabi identities, and hence would distance them from
Punjabiness. Consequently, neither of the two linguistic
choices available is completely acceptable, because they
both oppose values that the expatriate-raised Punjabis
identify with. The best compromise is the use of risk-free
forms of Punjabi, i.e., common borrowings from Punjabi and
fixed phrases often repeated in the Punjabi community.
Creative use of Punjabi by more fluent posters is tolerated
to the extent that the passive bilingual knowledge of non-
fluent speakers allows them to understand the posts. Fluent
speakers, in turn, limit their use of Punjabi in keeping
with the increasingly marginalized status of Punjabi in the
expatriate community as a whole.
The Prestige Status of English in South Asia
Another factor that promotes use of English among
first-generation Punjabis is the special status enjoyed by
the English language throughout South Asia. English was
introduced to India under the British colonial
administration of India, which then included what is now
Pakistan and Bangladesh. From colonial times on, English
has been widely used in India as the dominant language of
government and higher education (Fasold, 1984). Among
educated people, English is preferred for inter-ethnic
communication, since it has no significant population of
native speakers, and is therefore ethnically neutral, an
important asset in a society with many different ethno-
linguistic identities within it. Symbolically, English
confers high status to those who can use it, and it has
strong associations with "modernity" (Pandharipande, 1990).
Even those who have no formal education make every possible
effort to learn and use whatever English they can.
Recent Punjabi expatriates participating on
soc.culture.punjab are typically individuals who, by reasons
of social class, education, etc., have had substantial
exposure to English. Even if English is not the most fluent
language for these people, it is the one that is most
prestigious for them to use, especially in contexts where
others are also using English. The fact that many recent
expatriates are studying abroad in English-speaking
countries reinforces their disposition toward English. In
order to represent themselves as (foreign-)educated and
respectworthy, they must use English, even among their
Punjabi compatriots. The situation on Usenet is thus
similar to that of other forms of discourse among educated
people in India (Fasold, 1984; Pandharipande, 1990, 1992)
and in post-colonial societies more generally (Jacobson,
1990, Myers-Scotton, 1993a).
Usenet Language Norms
Finally, an important factor conditioning the choice of
language on soc.culture.punjab is the fact that it occurs on
Usenet. Usenet includes over 10,000 newsgroups which are
regularly carried by most hosts, and an indeterminate number
of others not regularly carried, or carried only by certain
hosts, in non-English speaking countries (e.g., a large
number of German newsgroups are not carried in the newsfeed
of most US universities). However, among the groups most
commonly carried, only a small minority carry newsfeed in
languages other than English (e.g., the fj groups in
Japanese, and some newsgroups in soc.culture pertaining to
Spanish and French-speaking countries). Although the number
of such groups appears to be increasing, it is still small
in comparison to the majority of Usenet newsfeed. In other
words, the "virtual" linguistic environment in which
soc.culture.punjab is situated is one which is dominated by
English. The poster above who commented that "English is
the best way to get my message across" was clearly
referencing himself to a language choice norm for Usenet
The influence of Usenet communication norms on
soc.culture.punjab is not merely felt in the predominance of
English. Other Usenet norms, influencing both message form
and content, are evident as well. Both the incidence and
nature of flaming on soc.culture.punjab are remarkably like
those on many other Usenet groups (cf. Kim & Raja, 1991).
Misogyny and feminist-baiting also occur on soc.culture.
punjab in ways typical of Usenet (cf. Sutton, 1994).
Likewise one finds frequent allusions to programming in the
computer language C (e.g., a joke written in C), whose
development and spread is intricately interwoven with Unix
and Usenet. All three of these behaviors are especially
common on Usenet, as well as being characteristic of the
Internet more generally. As regards these behaviors,
Punjabi posters are virtually indistinguishable from native
English-speaking posters in the US.
Thus we must count four factors influencing the
relative distribution of English and Punjabi on
soc.culture.punjab: inter-generational language shift in
expatriate Punjabi communities, cultural ambivalence toward
Punjabi, the prestige status of English in the Punjab and
South Asia more generally, and Internet norms of
communication. The result of the combined influence of
these four factors is a very limited use of Punjabi on the
newsgroup, despite the fact that all or most of the
participants are ethnic Punjabis. At the same time,
strategic and selective use of Punjabi serves to signal
Punjabi identity. As a consequence, enough Punjabi
(especially borrowed words) is used that it would be
difficult to follow the discourse on soc.culture.punjab
without at least some knowledge of Punjabi. However,
Punjabi words are embedded in a kind of codeswitching that
heavily favors English, because of its prestige and inter-
ethnic value, and because of its associations with the
modern technological environment of Usenet.
The findings presented here have implications for the
future use of Punjabi on soc.culture.punjab, as well as for
the use of other non-English languages on the Internet more
generally. Given that the patterns of choice observed on
soc.culture.punjab are reflective of a general pattern of
language shift in immigrant communities, we might predict
that further expression of this tendency would lead to even
greater marginalization of Punjabi language use. There is
some evidence to support this view. As of July 1996, the
frequency of Punjabi use on soc.culture.punjab was down to
about 3% of tokens (as compared to 15% in the earlier
sample). Since there has been a large turnover among the
posters on the newsgroup in the last two years, it is
difficult to interpret this finding conclusively, but the
indication is that Punjabi use is shifting increasingly to
English. Continued monitoring of soc.culture.punjab will
help to ascertain whether this observation represents a
To the extent that the sociolinguistic conditions
elsewhere are similar to those influencing soc.culture.
punjab, we might expect to see similar patterns of language
choice on other Usenet newsgroups. In many former British
colonies, English has a status similar to that found in
India. Moreover, inter-generational language shift is
common among many expatriate communities in the US, UK and
Canada. Other factors are relatively constant; for example,
the predominance of English on Usenet will not change
appreciably without a large increase in the number of
non-English posters or commonly carried groups. These
factors, individually and taken together, are predicted to
favor the use of English over other languages on-line.
Conversely, a relative absence of English influence in
the native culture, and the existence of users logging on
regularly from a country where their native language is
dominant and high prestige, could have the opposite effect,
that of favoring use of the native language on-line. There
are a number of Usenet newsgroups in which some other
language is used more than Punjabi is used on soc.culture.
punjab (e.g., those involving German, Spanish, French,
Finnish, Japanese, and Mandarin Chinese). Clearly there is
appreciable variation in the use of non-English languages on
the Internet, including possibly the effects of additional
factors not discussed here, for example, whether the
language uses a roman or non-roman writing system, and
whether the mode of CMC is synchronous or asynchronous. Our
understanding of the future shape of the largest language
contact situation in known history, and of the dynamics of
the present that mould it, depends on our understanding of
such factors. Thus it is important that further questions
be asked about language choice in other Usenet newsgroups,
in IRC, and in other modes of CMC, as language contact in
multilingual, multi-ethnic cyberspace evolves.
 Posters' countries of residence are normally
evident from their e-mail addresses. A poster's length and
time of residence in a country is often known in one of two
ways. In the early days of the newsgroup, a number of
posters posted personal biographical information, often
including birthplace, places they have resided, and
approximate dates. Alternatively, in the context of a
discussion, a poster might also give the same information in
relating a personal experience or attitude.
 Translation of this passage by Harbans Singh
(_Heritage of the Sikhs_).
 Punjabi poetry tends to be highly stylized, non-
colloquial and difficult to understand, since the Punjabi
poetic tradition draws from several literary traditions of
 The only exception to this is the single Punjabi
message of a female poster who otherwise posts in English.
Her only Punjabi post was an insult (directed at "Hindustani
Punjabis") without any prior provocation. She later posted
a message acknowledging the inappropriateness of her Punjabi
 Recent expatriates appear to place a higher value
on the maintenance of Punjabi language and culture, and
sometimes deride and stigmatize members of the community who
do not. Second-generation members, known to the South Asian
community as "ABCD" -- American Born Confused Desi -- suffer
this stigma, which compounds their insecurity about not
knowing the appropriate language or cultural norms necessary
to function fully in the community.
Eastman, C. (1992). Codeswitching as an urban language
contact phenomenon. Journal of Multilingual and
Multicultural Development, 13 (1), 1-17.
Fasold, R. (1984). Sociolinguistics of society. Oxford:
Gibson, M. (1988). Accommodation without assimilation.
Ithaca NY: Cornell University Press.
Hill, J. (1993). Structure and practice in language shift.
In K. Hyltenstam & A. Viberg (Eds.), Progression and
Regression in Language (pp. 68-93). Cambridge, UK:
Cambridge University Press.
Jacobson, R. (Ed.) (1990). Codeswitching as a worldwide
phenomenon. New York: Peter Lang.
Jacobson, R. (1990). Introduction. In R. Jacobson (Ed.),
Jones, S.G. (Ed.) (1995a). Cybersociety. Thousand Oaks,
Jones, S.G. (1995b). Understanding community in the
information age. In S. Jones (Ed.), pp. 10-35.
Kim, M-S, & Raja, N.S. (1991). Verbal aggression and
self-disclosure on computer bulletin boards.
Washington, D.C.: ERIC Clearinghouse on Languages and
MacKinnon, R.C. (1995). Searching for the Leviathan in
Usenet. In S. Jones (Ed.), pp. 112-137.
McLaughlin, M.L., Osborne, K.K., & Smith, C.B. (1995).
Standards of conduct on Usenet. In S. Jones (Ed.), pp.
Myers-Scotton, C. (1993a). Social motivations for
codeswitching. Oxford: Oxford University Press.
Myers-Scotton, C. (1993b). Duelling languages. Oxford:
Oxford University Press.
Pandharipande, R. (1990). Formal and functional constraints
on code-mixing. In R. Jacobson (Ed.), pp. 15-31.
Pandharipande, R. (1992). Language shift in India, issues
and implications. In W. Fase, K. Jaspaert & S. Kroon
(Eds.), Maintenance and loss of minority languages (pp.
253-275). Amsterdam: John Benjamins.
Rheingold, H. (1993). Virtual community: Homesteading on
the electronic frontier. Reading, MA: Addison-
Singh, H. (1994). Heritage of the Sikhs (2nd ed.). New
Sutton, L. (1994). Using Usenet: Gender, power and silence
in electronic discourse. Proceedings of the 20th
Annual Meeting of the Berkeley Linguistics Society.
Berkeley, CA: Berkeley Linguistics Society.
Tesitelova, M. (1992). Quantitative linguistics.
Amsterdam: John Benjamins.
Communication Institute for Online Scholarship, Inc.
This file may not be publicly distributed or reproduced
without written permission of the Communication Institute
for Online Scholarship, P.O. Box 57, Rotterdam Jct., NY
12150 USA (phone: 518-887-2443).