Language and variation

Text-only Preview

Chapter 7: Language variation
Chapter 7

Language and variation

7.1 Introduction

The study of language variation is an important part of sociolinguistics, to the extent
that it requires reference to social factors. Languages vary from one place to another,
from one social group to another, and from one situation to another, and these are the
main topics of this chapter. They are treated in § 6.2 Geographical variation, § 6.3
Social variation, and § 6.4 Contextual variation, respectively. Language variation is a
political issue, which is treated in § 6.5 Language policy in Asia and Africa.

7.1.1 Linguistic items and varieties

We talk and write about languages, dialects, sociolects, accents, jargons, registers,
and so on and so forth, but none of these terms can be taken for granted and many of
them are difficult to define in a satisfactory way.
For example, if dialect is defined as a geographical subdivision of a language,
we do not come very far without a definition of language, and subdivision is not a
very clear concept, either. As the British linguist Richard Hudson writes, the
«discussion will be easier if we have some technical terms to use, as we need to
distance ourselves somewhat from the concepts represented by the words language
and dialect, which are a reasonable reflection of our lay culture, called ‘common-
sense knowledge’ […], but not helpful in sociolinguistics.»1
Therefore, we need to start our discussion with a terminology that does not take
very much for granted. Our most basic terms will be linguistic item and variety. LINGUISTIC ITEM

Sociolinguists in most cases study social distribution of particular linguistic items,
for example words, sounds, or grammatical constructions. Let us give some examples
of linguistic items.
The English pronouns yous ‘2nd person plural’ and you ‘2nd person singular or
plural’ are linguistic items, and they have different social distributions; the former is
found in certain non-standard varieties of English, while the latter occurs inter alia in
all standard varieties and some other non-standard varieties.

1 Page 21 in Richard Hudson 1996: Sociolinguistics. 2nd ed. (Cambridge Textbooks in Linguistics.)
Cambridge: Cambridge University Press.

Chapter 7: Language variation
If you take a look at a dialect atlas of England,2 you will find that roughly, the word
child is used in southern England and in Midland, while bairn is used in northern
England. Child and bairn are different linguistic items.
In England, the sound /√/, as in sun /»s√n/, is a typical southern sound, found in
southern England and in South Midland, while this sound is not used among speakers
of dialects in North Midland and northern England, where, for example, the word sun
is pronounced /»sUn/, with the sound /U/, which is found in put /»pUt/ in most dialects
also in the South (some areas have /√/). The English phonemes /√/ and /U/ are
different linguistic items.
The suffix –ing of written English, as in coming, is pronounced /IN/ and /In/, as
in /»k√mIN/ and /»k√mIn/, and the two pronunciations have different social
distributions: the former is a typical standard pronunciation and the latter a typical
non-standard pronunciation. The English suffixes /IN/ and /In/ are different linguistic
In the English dialects of England, the most widespread past tense of catch
/»kQtS/ is catched /»kQtSt/, while the standard dialect and some other dialects have
caught /»kç˘t/. The English past tense forms catched /»kQtSt/ and caught /»kç˘t/ are
different linguistic items.
Give it to me! is an example of a Standard English grammatical construction,
with a verb (here in the imperative) followed by a pronoun referring to a patient and a
prepositional phrase containing the preposition to plus a pronoun referring to a
recipient. In traditional dialects of England, this construction is not very common,
being found primarily in the South-west and in some areas on the south-eastern coast
(including the London area). The construction with the widest geographical
distribution in England is Give me it! and Give it me! These three sentences, Give it to
, Give me it!, and Give it me!, are instances of three different grammatical
constructions, each of which is a linguistic item. VARIETY

There are many ways of speaking, and each way of speaking is a variety. In a more
precise manner, a variety may be defined as a set of linguistic items with similar
social distribution

It should be emphasized that a variety is not necessarily a «full-fledged
language», with a large vocabulary and grammar. It may simply be a small set of
linguistic items, as is the case with a slang, which may typically be defined as a quite
restricted set of new words and new meanings of older words, mixed with linguistic
items with a much larger social distribution; cf. § 6.3.2. and (1g) on the next page.
In (1), we have given some examples of sentences in different varieties of
language. On the basis of these examples, we can ask some of the central questions of
this chapter, like: Do these varieties represent the same or different languages? Do
these varieties represent the same or different dialects of the same language?
– and so
on. More concretely, one could for example ask how many different languages are
represented in (1), and there is no unique answer.

2 A very good one is Clive Upton and J. D. A. Widdowson 1996: An Atlas of English dialects. Oxford:
Oxford University Press. Most of the examples in § are taken from this book.
3 Cf. Richard Hudson (ibid.).

Chapter 7: Language variation
(1) Varieties of language
Standard English. No one has gone to the post office yet.
Jamaican Creole .Nobadi no gaan a puos yet. ‘No one has gone to the post office yet.’
Southern US white Non-Standard dialect from Atlanta. Nobody don’t like a boss hardly. ‘Hardly
anybody likes a boss.’
New Guinea Pidgin (Tok Pisin). Papa, min bin mekim sin long God na long yu. ‘Father, I have
sinned against God and against you.’
Older Standard English of the ‘King James version’ Bible.Father, I have sinned against heaven,
and in thy sight.
Scots, from Leith. When ah wis a boy ma mither an faither died. ‘When I was a boy my mother
and father died.’
Standard English & English slang (ball-ache) Walking 5 miles to work is a real ball-ache.
‘Walking 5 miles to work is really inconvenient.’
Chadian Spoken Arabic of Ulâd Eli.’ Amm Muusa daxalat zeribt al-bagar. ‘Mûsa’s mother
entered the enclosure of the cows.’
Moroccan Spoken Arabic.B!i˘t n©kri sayyara lmudd©t usbu˘¿. ‘I would like to hire a car for a
Standard Maltese.Mart is-sultan marida �afna. ‘The sultan’s wife is very ill.’
Standard Written Arabic.Ra&aytu na\san !ayra sukka\ni Makkata. ‘I saw people who were not the
inhabitants of Mecca.’

The varieties in (1a) – (1g) may all be called English is some sense, but it is not at
evident that these varieties represent the same language. Likewise, the varieties in
(1h) – (1k) may be characterized as Arabic, but this does not necessarily mean that
only one language is involved, and what we have referred to here as one variety,
Standard Written Arabic in (1p), may be divided into at least two different varieties,
Classical Arabic (the language of the Qur&a\n and writers like Ibn Khaldu\n) and
Modern Literary Arabic (the language of modern newspapers and many modern

7.2 Geographical variation: language and dialect

We often talk about the Chinese language, the Hindi language, the Arabic language,
and the Fula language, without thinking about how problematic these terms are. Of
course, they are not at all meaningless, but their meanings are often rather different
from what is generally assumed.

7.2.1 Fula: A dialect continuum

We shall take a look at Fula language, which is spoken in 17 countries, most of them
in West Africa, especially in Sahel, the savanna belt south of the Sahara desert, from
Mauritania and Senegal in the west, through Guinea, Mali, Burkina Faso, Niger,
Nigeria, Cameroon, Chad, and neighboring countries areas; cf. MAP 1.4 The Fula
speaking area is not geographically continuous; there are many Fula speaking areas
across Sahel, interrupted by areas where hundreds of other languages are spoken.

4 There may be close to 20 million first language speakers of Fula and more than 7 million second
language speakers.

Chapter 7: Language variation

MAP 1. The main Fulani of West Africa are dark (blue) on the map.

7.2.2 Dialect

It is generally assumed that Fula is a language, that is, a single language, with a
number of dialects. In this sense, a dialect is regarded as a geographical variety of a
language, spoken in a certain area, and being different in some linguistic items from
other geographical varieties of the same language.
This definition of dialect is in common use among linguists, and differs from a
usage found in several European language communities among non-linguists, where
dialect is often used about «provincial» varieties that differ from the standard dia-
, which is then regarded as the «proper language»; we shall come back to the
standard dialect in § 6.2.6. The standard dialect is then regarded as the «non-dialectal»
variety of the language.
The dialect definition presented above is problematic in an important way: It
presupposes a satisfactory definition of a language, but such a definition does not
exist. We shall come back to this problem in the following.

7.2.3 Dialect continuum

The different Fula speaking areas in West Africa may be referred to as dialect areas,
and there are between ten and fifteen major dialect areas. The most important ones are
found in (i) northern Senegal and southern Mauritania, (2) Guinea, (3) Mali, (4)
Burkina Faso, western Nigeria, and western Niger, (5) central Nigeria, and (6) eastern
Nigeria and northern Cameroon. When speakers from neighboring dialect areas meet,
they can communicate with each other without problems, each one her or his native
variety. However, if a Fula speaker from one end of West Africa meets a Fula speaker
from the other end, there may be problems, although communication is still possible,
perhaps with some exceptions (speakers from eastern Nigeria and northern Cameroon

Chapter 7: Language variation
would have difficulties in understanding a speaker from Guinea), and depending to a
certain extent upon individual speakers—everybody is not equally good at
understanding dialects differing from their own.5 Undoubtedly, the Fula speaking
areas of West Africa can be described as a dialect continuum, which may be defined
as follows:6

A dialect continuum is a chain of dialects, let us say dialects 1–10, with the
following property: Speakers of dialect 1 understand dialect 2 extremely well.
Speakers of dialect 1 and dialect 3 understand each other rather less well, and
speakers of dialect 1 and dialect 4 less well again. There comes to a point,
however, say at dialect 5, where dialect 1 is no longer intelligible to the local
people and vice versa.

There are many dialect continua around the world. For example, the rural dialects of
Portugal, Spain, France, and Italy constitute a dialect continuum. There is always
mutual intelligibility between the dialects of neighboring villages, throughout the
area, although the intelligibility decreases as the distance increases. Another European
dialect continuum is found in Scandinavia, where there is always intelligibility among
neighboring dialects of the North Germanic language area of Denmark, Sweden, and
Norway. The Arabic dialects from Morocco to Iraq also constitute a dialect
continuum, and so does also a large part of the Indo-Aryan language area of northern

Let us go back to the definition of dialect for a minute. One might try to
formulate a new definition on the basis of the dialect continuum phenomenon, instead
of on the basis of a language, proposing that a dialect is one of several mutually
intelligible geographical varieties. Now, there are many «languages» that are mutually
intelligible, for example Danish, Swedish, and Norwegian, or several of the Turkic
languages, including Turkish and its closest neighbors, and our new definition of
dialect forces us to claim that these «languages» are in a «dialect relationship» to each
other. This may be quite meaningful, because, as we shall come back to, «languages»
may emerge on several places along a dialect continuum, and the number may even
vary at different times.

7.2.4 Isoglosses

What makes dialects—as well as languages—different is their differing sets of
linguistic items. Take a look at TABLE 1 (next page) with some words from three
different Fula dialects, Fuuta Tooro (Senegal / Mauritania), Maasina (Mali), and
Aadamaawa (Nigeria / Cameroon).
The two first words in TABLE 1, those meaning ‘land’ and ‘book’, shows a
case where the geographically central dialect of Maasina has different forms from the
geographically peripheral dialects of Fuuta Tooro and Aadamaawa. This indicates that
an innovation has taken place in the Maasina dialect, which has changed the fricatives
/f/ and /s/ (the only fricatives in most dialects of Fula) into /w/ and /j/ (written y) in
coda position (cf. § 2.4.5 Syllables). It is possible to show on a map the geographical
distribution of those varieties of Fula that has fricatives in coda position and those

5 The author of this chapter has personal experiences in this respect. First, he learnt the Fula dialect of
eastern Nigeria and northern Cameroon. Some years later he went to Mali and experienced that he
could communicate quite well with Fula speakers there. The main problems were at the lexical level.
6 Adopted from William Downes 1998: Language and Society. 2nd ed. Cambridge: Cambridge
University Press.

Chapter 7: Language variation
varieties that do not, and the border between the former varieties and the latter
varieties is referred to as an isogloss.

‘to write’
TABLE 1. Words from three Fula dialects

Another dialectal difference in Fula is the one
between those dialects where the infinitive suffix is
and those where it is –go. Here we can observe that
Aadamaawa differs from Maasina and Fuuta Tooro.
Consequently, there is an isogloss between
Aadamaawa and the two others. Historically,
Aadamaawa is the innovating dialect.

Finally, Fuuta Tooro differs from Maasina and
Aadamaawa by having the active habitual suffix –at,
while the two others have –an. There is an isogloss
separating Fuuta Tooro from the two other dialects.
Historically, Maasina and Aadamaawa have
innovated: they do not allow word final suffixes
nding in /t/ any more.

In FIG. 1, we have drawn a schematic map
PHOTO 1. A Fulani shepherd
showing the mentioned isoglosses and how they

separate the three dialects.

Fuuta Tooro



Varieties inside

Varieties inside

Varieties inside

this isogloss do

this isogloss do

this isogloss
not have
not have word
have the

fricatives in the
final suffixes

infinitive suffix
ending in /t/.

FIG. 1. Fula isoglosses

7.2.5 Abstand languages and Ausbau languages

The German sociolinguistic Heinz Kloss has introduced the fruitful distinction
between Abstand languages or languages by distance and Ausbau languages or
languages by extension, which has been further elaborated by the British sociolinguist
Peter Trudgill.7 Let us explain the difference between them with some examples.

7 Cf. for example Peter Trudgill 1997: Norwegian as a normal language. P. 151–158 in Unn Røyneland
(ed.): Language contact and language conflict. Volda: Volda College.

Chapter 7: Language variation KOREAN: AN ABSTAND

Everybody regards Korean as a language.
It is relatively homogeneous, with good
mutual intelligibility among speakers from
different areas.8 Furthermore, the Korean
dialects could never reasonably be regard-
ed as dialects of any other language; it is
not even genetically related to any of its
immediate neighbors. Many scholars re-
gard Korean and Japanese as being genet-
ically related, but the their relationship is
very distant. They are much more distant
than for example English and French.

Korean can be regarded as a inde-
pendent language because it is a ‘language
by distance’. It is a relatively homogene-
ous dialect cluster that is only distantly
related to other languages. There is also

only one Korean written standard

language. There are some minor ortho-
MAP 2. Korea
graphic divergences between North and

South Korea that have arisen since the division in 1945, but the spellings are not
diverse enough to justify talking about two different languages. Cf. § 6.5.2. TURKIC AUSBAU LANGUAGES

MAP 3. Central Asia.

8 Cf. Ho-Min Sohn 1999: The Korean Language. (Cambridge Language Surveys.) Cambridge:
Cambridge University Press.

Chapter 7: Language variation
As mentioned in Ch. 4, Language families, the Turkic language family has two main
branches, r-Turkic and z-Turkic (or Common Turkic). Among the z-Turkic
languages are Turkmen, Uzbek, Kazakh, Kirgiz, and Uighur, which are spoken
across a huge continuous area in Central Asia; cf. MAP 3. In the introduction of
Central Asia Phrasebook,9 we are told:

As one travels throughout the Turkic heart and periphery of Central Asia it becomes abundantly
clear that the Turkic languages are really dialects of one another and not distinct languages at all.
This means that in learning one, all the rest are understandable to a large degree. Herein lies the fun
and the challenge.

To illustrated how close these languages are, we have included TABLE 2, which shows
the numbers from 1 to 10 in Turkish and the five languages mentioned above.10 As a
speaker of for example Norwegian will notice, the numerals do not differ more
between these languages than numerals in different Norwegian dialects.


TABLE 2. The numerals from 1 to 10 in some Turkic languages.

These Turkic languages constitute a continuum of closely related dialects. If the
political situation had been different, the number of languages could also have been
different. As Peter Trudgill (1997 : 152) writes, most languages, «have arisen out of
dialect continuum situations and are considered separate languages for reasons that
are by no means entirely linguistic». These are the typical Ausbau languages.

Counting ‘languages’ in the Turkic dialect continuum implies more or less
counting standard languages, which are treated in the next paragraph.

7.2.6 Standard languages

The adjective standard means ‘recognized as correct or acceptable’, and a standard
is a variety that in different ways is recognized as more correct and accept-
able than other varieties. In many ways, standard variety is an equally appropriate
designation. It has the following prototypical properties:

9 Justin Jon Rudelson 1998: Central Asia Phrasebook. Languages of the Silk Road. Hawthorn: Lonely
10 The orthographies of the individual languages have been adjusted to that of Turkish, to make the
similarities between them even clearer.

Chapter 7: Language variation
• It is the variety of used by educated users, e.g. those in the professions, the
media, and so on.
• It is the variety defined in dictionaries, grammars, and usage guides.
• It is regarded as more correct and socially acceptable than other varieties.
• It enjoys greater prestige than dialects and non-standard varieties: non-
standard varieties are felt to be the province of the less educated.
• It is used as a written language.
• It is used in important functions in the society – in the government, the
parliament, courts, bureaucracy, education, literature, trade, and industry.

The written language plays an important role in this connection. In general, only
standard languages have a stable written form, which is what is general taught in

It is very important to be aware of the fact that not all languages have a
standard variety, and that standard varieties rise and fall. The reverse of standardi-
zation is dialectization. In a European context, Scots, Low German, and Occitan /
Provençal have been standard languages earlier, but they are no longer, and tend to be
regarded as subordinate to English, High German, and French, respectively.

7.3 Social variation: sociolect and slang/jargon

It should have come as no surprise that language varies geographically. We are not
surprised to hear that people who live far from each other speak more differently than
people who live close to each other, because those who live close to each other have
more contact with each other than those who live far away from each other.

The terms closeness and distance originally come from the spatial domain (cf.
§ 2.2.1 Domains), but metaphorically they have been transferred to the social domain.
For example, we talk about the varying social distance between individuals in a
society, and differences in social distance correlate strongly with language variation.

7.3.1 Social organization

The organization of society can be approached from to opposing angles, the angles of
social network and social stratification. Social stratification concerns the
hierarchical structure of a society, arising from inequalities of wealth and power. On
the other hand, social network concerns the dimensions of solidarity between
individuals in their everyday contacts. SOCIAL NETWORK

An individual is a part of a social network and has stronger and looser ties with other
individuals. Networks vary in strength, which primarily is based upon density and
multiplexity, which are defined by Milroy and Milroy (1997 : 60) in the following

A maximally dense network is one in which everyone knows everyone else, and a multiplex
relationship is one in which A interacts with B in more than one capacity (for example, as
workmate and friend).

11 Milroy and Milroy (1997): cf. the preceding footnote.

Chapter 7: Language variation

When networks are strong, language tends to change more slowly, and stigmatized
and low-status language items persist. We shall come back to this issue in § 6.3.2. SOCIAL STRATIFICATION

There are hierarchical social structures in most or all societies, but not everywhere of
the same type. In large parts of Europe, for example, society started to change about
200 years ago from a hierarchy of rank or station to a hierarchy of class. In the rank
society, people are born with a certain rank, and there is low social mobility. In the
class society, people are also born into a certain class, but there is a high social
These differences influence the language situation in a society. In the rank
society language differences do not play an important role. People learn to speak the
variety they hear in their social network, and continue to speak that way for the rest of
their lives. Their rank is primarily determined by their family background, and people
cannot change their social status by changing their language. Highest in the hierarchy
you often find an aristocracy
In the class society things are quite different. Many people climb «upwards»
in the class hierarchy, and as a part of their effort to change their social status, they
change their language in the direction of people higher up in the hierarchy.

7.3.2 Sociolect

In the traditional European rank society people generally spoke the dialect of their
home area, and there was only minor variation between the ranks. On the basis of a
person’s language variety you could easily locate her or him geographically, but not at
all to the same degree socially.

In the end of the 18th century and in the beginning of the 19th, this society
started to change, as a consequence of industrialization, which created new social
strata—particularly a working class and a bourgeoisie or middle class—and oppor-
tunities for people to improve their economical and social status.
In the book ‘Talking Proper’. The Rise of Accent as Social Symbol, Lynda
Mugglestone tells the story about what happened in England.12 For centuries England
had had a standard written language, but no standard spoken language. In the end of
the 18th century, however, this situation started to change drastically. The middle class
consciously changed their speaking habits in the direction of the most prestigious
variety of spoken English, which was the variety used at the royal court in London. At
first, this created a situation where the upper class (the aristocracy)—who evidently
could not improve their social status by any means—and the lower class (the working
class) spoke the local dialect, while the middle class adopted the new spoken
standard, which varied much less from place to place.
The middle class changed their language habits more than the upper class and
the lower class not only due to their desire to rise socially, but also because of net-
work differences. As Milroy and Milroy (ibid. : 61) writes,

close-knit solidarity ties are characteristic of lower and higher social groups, and […] in the middle
sectors of society, social network density and multiplexity tend to be weak. […] [D]ifferent kinds of
social network structure do not occur accidentally, but “fall out” naturally from different life-

12 Lynda Mugglestone 1995: ‘Talking Proper’. The Rise of Accent as Social Symbol. Oxford :
Clarendon Press.