Blom and gumperz 1972 social meaning in linguistic structures

SOCIAL MEANING IN LINGUISTIC STRUCTURE: CODE-SWITCHING IN NORWAY 219

Irfinitive Present Past Past Participle

(N (B) (R) (B) (R) (B) (R) (B)

finj finne finj finner Janj Jant Janje Junnet (find) raTa or va vcere e eer va var vore vcert (be)

jeer jore jeer jar jol jore jort jort (do)

lees lese lees leser lcest leste leest lest (read)

Other important dialect features appear in pronouns, common adverbs of time, place, and manner, conjunctions, and other grammatically significant function words. Here is a list of some of the most common distinctive forms of personal pronouns and possessive pronouns:

(B) (R)

jeejj °a (I)

meejj mea (me) deejj dea (you) hann hanj (he)

(B) (R) hunn ho (she) hanns hanjs (his) hennes hinjers (hers) dere dakk (you)(plural) di deemm* (theirs)

*' Sometimes also di and deres. [ ... ]

These data constitute empirical evidence to support the view of the dialect as a /ilkH,nct linguistic entity.

[ ... ]

HAI\lAIVIP,L AND BOKMAL AS CODES IN A REPERTOIRE

[ .. ,] illustrate the differences between Ranamal and Bokmal] in a sentence pair like hanj bar nilsen's pagsonat and its Bokmal equivalent hann bar pa nilsen's pagsonat "He lives in

's pensionat," only the realizations of lal, I'll, and Injl which appear in our list of '.~".

·b.etw"entwo clearly distingUishable articulation points; rather, the shift takes the form of 'elisf,lae:en}erlt along a scale in which palatalized consonants show at least three degrees

p.l.tal'izatiOl), strong [nil, weak [ni], and zero [n] and (a( and h£! each show three of retraction and lowering.

a switch from Nonvegian to English implies a shift between two distinct rue'tur'al wholes, the Bokmal-Ranamal alternation, in phonology at least, seems more

220 JAN-PETEA BLOM AND JOHN J. GUMPERZ

similar to conditions described by Labov (1966b) for New York speech. A speaker's standard and dialect performance can be accounted for by a single phonetic system.

[ ... J

The effect of structural similarities on speakers' perception of speech differences is somewhat counterbalanced by the fact that choice among these variables is always restricted by sociolinguistic selection constraints [also referred to as co-occurrence rules1 such that if, for instance, a person selects a standard morphological variant in one part of an utterance, this first choice also implies selection of pronunciation variables tending toward the standard end of the scale.

[ ... J

The most reasonable assumption is that the linguistic separateness between the dialect and the standard, i.e., the maintenance of distinct alternates for common inflectional morphemes and function, is conditioned by social factors.

Some idea of how this came about can be obtained by considering the conditions under which the two varieties are learned. The dialect is acquired in most homes and in the sphere of domestic and friendship relations. As a result, it has acquired the flavor of these locally based relationships. However, dialect speakers learn the standard in school and in church, at a time when they are also introduced to national Norwegian values. It has therefore become associated with such pan-Norwegian activity systems.

Since the adult population has equal access to both sets of variants, however, the developmental argument does not provide sufficient explanation for the maintenance of distinctness. Immigrants to urban centers around the world, e.g., frequently give up their languages after a generation if social conditions are favorable to language shift. The hypothesis suggests itself, therefore, that given the initial acquisition patterns, the dialect and the standard remain separate because of the cultural identities they communicate and the social values implied therein. It is this aspect of the problem that we intend to explore in the remaining sections of the article.

[ ... J

Effective communication requires that speakers and audiences agree both on the meaning of words and on the social import or values attached to choice of expression. Our discussions will be confined to the latter. We will use the term social Significance, or social meaning, to refer to the social value implied when an utterance is used in a certain context.

In general, the assignment ot value to particular objects or acts is as arbitrary as the referential naming of objects. Just as a particular term may refer to a round object in one group and a square object in another, so also the value of actions or utterances may vary. Thus the same term may indicate geographical distinctions in one community and symbolize social stratification elsewhere. Social meanings differ from referential mean- ings in the way in which they are coded. Whereas reference is coded largely through words, social meaning can attach not only to acoustic signs but also to settings, to items of background knowledge, as well as to particular word sequences. In Hemnes, values attached to a person's family background or to his reputation as a fisherman

SOCIAL MEANING IN LINGUISTIC STRUCTURE: CODE*SWITCHING IN NORWAY 221

important in understanding what he says and influence the selection of responses to his

actions.

[ •.. J

LOCAL ORGANIZATION AND VALUES

[ ... J The majority of those who claim local descent show a strong sense of local identification. To be a hcemnesvcerin8 "Hemnes resident" in their view is like belonging to a team characterized by commonalty of descent. [ ... J The dialect is an important marker of their common culture.

[ ... J

The meaning attached to local descent and dialect use-to being part of the "local team"-is clearly seen when we consider those members of the community who dissociate themselves from this "team." Traditionally, in northern Norway the local community of equals was separated from the landowning commercial and administrative elite by a wide gulf of social and judicial inequality. Since the latter were the introducers and users of standard Norwegian, the standard form was-and to some extent still is-associated with this inequality of status. Many of the functions of the former elite have now been incorporated into the local social system. Individuals who fill these functions, however, continue to be largely of nonlocal descent. Although they may pay lip service to locally accepted rules of etiquette and use the dialect on occasion, their experience elsewhere in Norway, where differences in education, influence, and prestige

much more pronounced, leads them to associate the dialect with lack of education and sophistication. Therefore, they show a clear preference for the standard.

Such attitudes are unacceptable to locals, who view lack of respect for and refusal speak the dialect as an expression of social distance and contempt for the "local team"

and its community spirit. It is not surprising, therefore, that their loyalty to the dialect is. thereby reaffirmed. For a local resident to employ (B) forms with other local residents

in their view to snakk fint or to snakk jalat - "to put on airs. ')

[ ... J

CONSTRAINTS

.. 1 There is by no means a simple one-to-one relationship between specific speech and specific social identities. Apart from the fact that values attached to

.::I'la~:e usage vary with social background, the same individual need not be absolutely ,consiSiterLt in all his actions. He may wish to appear as a member of the local team on

occasions, while identifying with middle-class values on others. In order to determine the social Significance of anyone utterance, we need additional information

the contextual clues by which natives arrive at correct interpretations of social !il1eaning.

[ ... J

222 JAN·PETER BLOM AND JOHN J. GUMPERZ

We will use the term setting to indicate the y\lay in which natives claSSify their ecological environment into distinct locales [e.g. the home, workshops and plants, the school, etc.].

[ ... ]

A closer specification of social constraints is possible if we concentrate on activities carried on by particular constellations of personnel, gathered in particular settings during a particular span of time [e.g. class sessions as opposed to meetings in the school]. We will use the term social situation to refer to these. [ ... ]

Thus alternative social definitions of the situation may occur within the same setting, depending on the opportunities and constraints on interaction offered by a shift in personnel and/or object of the interaction. Such definitions always manifest themselves in what we would prefer to call a social event. Events center around one or at the most a limited range of topics and are distinguishable because of their sequential structure. They are marked by stereotyped and thus recognizable opening and closing routines. The distinction between situation and event can be clarified if we consider the behavior of Hemnes residents who are sometimes seen in the community office, first transacting their business in an officially correct manner, and then turning to one of the clerks and asking him to step aside for a private chat. The nOrms which apply to the two kinds of interaction differ; the break between the two is clearly marked. Therefore, they constitute two distinct social events, although the personnel and the locale remain the same.

The terms setting, social situation, and social event as used here can be considered three successively more complex stages in the speaker's processing of contextual information. [ ... 1 To demonstrate how these factors influence language usage in Hemnesberget, we turn now to some examples drawn from participant observation.

The fact that the dialect reflects local values suggests that it symbolizes relationships based on shared identities with local culture. Casual observations and recording of free speech among locals in homes, workshops, and the various public meeting places where such relationships are assumed do indeed show that only the dialect is used there. However, statuses defined with respect to the superimposed national Norwegian system elicit the standard. Examples of these are church services, presentation of text material in school, reports, and announcements-but not necessarily informal public appeals or political speeches-at public meetings. Similarly, meetings with tourists or other strangers elicit the standard at least until the participants' identity becomes more clearly known.

SITUATIONAL AND METAPHORICAL SWITCHING

When within the same setting the participants' definition of the social event changes, this change may be signaled among others by linguistic clues. On one occasion, when we, as outsiders, stepped up to a group of locals engaged in conversation, our arrival caused a significant alteration in the casual posture of the group. Hands were removed from pockets and looks changed. Predictably, our remarks elicited a code switch marked simultaneously by a change in channel cues (Le., sentence speed, ryhthm, more hesitation pauses, etc.) and by a shift from (R) to (B) grammar. Similarly, teachers report that while

SOCIAL MEANING IN LINGUISTIC STRUCTURE; CODE·SWITCHING IN NORWAY 223

formal lectures-where interruptions are not encouraged-are delivered in (B), the speakers will shift to (R) when they want to encourage open and free discussion among students. Each of these examples involves clear changes in the participants' definition of each other's rights and obligation. We will use the term situational switchin8 to refer to this kind of a language shift.

The notion of situational switching assumes a direct relationship between language and the social situation. The linguistic forms employed are critical features of the event in the sense that any violation of selection rules changes members' perception of the event. A person who uses the standard where only the dialect is appropriate violates commonly accepted norms. His action may terminate the conversation or bring about other social sanctions. To be sure, language choice is never completely determined; SOCiolinguistic variables must be investigated empirically. Furthermore, situations differ in the amount of freedom of choice allowed to speakers. Ritual events, like the well-known Vedic ceremonies of South Asia, constitute extreme examples of determination, where every care is taken to avoid even the slightest change in pronunciation or rhythm lest the effectiveness of the ceremony be destroyed. [ .. . J

In Hemnesberget, as our example will show later on, speakers are given a relatively \vide choice in vocabulary and some choice in syntax. Selection rules affect mainly the variables discussed previously. Values of these variables are sOciolinguistically determined in the sense that when, on the one hand, we speak of someone giving a classroom lecture or performing a Lutheran church service or talking to a tourist, we can safely assume that he is using (B) grammatical forms. On the other hand, two locals having a heart-to-heart talk will presumably speak in (R). If instead they are found speaking in (B), we conclude either that they do not identify with the values of the local team or that they are not having a heart-to-heart talk.

In contrast Vi'ith those instances where choice of variables is narrowly constrained by social norms, there are others in which participants are given considerably more latitude. Thus official community affairs are largely defined as nonlocal and hence the standard is appropriate. But since many individuals who carry out the relevant activities all know each other as fellow locals, they often interject casual statements in the dialect into their formal discussions. In the course of a morning spent at the community administration office, we noticed that clerks used both standard and dialect phrases, depending on whether they were talking about official affairs or not. Likewise, when residents step up to a clerk's desk, greeting and inquiries about family affairs tend to be exchanged in the dialect, while the business part of the transaction is carried on in the standard.

In neither of these cases is there any Significant change in definition of participants' mutual rights and obligations. The posture of speakers and channel clues of their speech remain the same. The language switch here relates to particular kinds of topics or subject matters rather than to change in social situation. Characteristically, the situations in question allow for the enactment of two or more different relationships among the same set of individuals. The choice of either (R) or (B) alludes to these relationships and thus generates meanings which are quite similar to those conveyed by the alternation between ty or '7 in the examples from Russian literature cited by Friedrich (1972). We will use the term metaphorical switchin8 for this phenomenon.

The semantic effect of metaphorical sVi'itching depends on the existence of regular relationships between variables and social situations of the type just discussed. The context in which one of a set of alternates is regularly used becomes part of its meaning,

224 JAN·PETER BLOM AND JOHN J. GUMPERZ

so that when this form is then employed in a context where it is not normal, it brings in some of the flavor of this original setting. Thus a phrase like "April is the cruelest month" is regarded as poetic because of its association with T. S. Eliot's poetry. When used in natural conversation, it gives that conversation some of the flavor of this poetry. Similarly, when (R) phrases are inserted metaphorically into a (B) conversation, this may, depending on the circumstances, add a special social meaning of confidentiality or privateness to the conversation.

The case of the local who, after finishing his business in the community office, turns to a clerk and asks him to step aside for a private chat further illustrates the contrast between metaphorical and role svvitching. By their constant alternation between the standard and the dialect during their business transaction, they alluded to the dual relationship which exists between them. The event was terminated when the local asked the clerk in the dialect whether he had time to step aside to talk about private affairs, suggesting in effect that they shift to a purely personal, local relationship. The clerk looked around and said, "Yes, we are not too busy." The two then stepped aside, although remaining in the same room, and their subsequent private discussion was appropriately carried on entirely in the dialect.

THE EXPERIMENT

Our discussion of verbal behavior so far has relied largely on deductive reasoning supported by unstructured ethnographic observation. Additional tests of our hypothesis are based on controlled text elicitation. We have stated that gatherings among friends and kin implying shared local identities must be carried on in the dialect. If we are correct in our hypothesis, then individuals involved in such friendly gatherings should not change speech variety regardless of whether they talk about local, national, or official matters.

In order to test this, we asked local acquaintances whom we knew to be part of the network of local relationships to arrange a friendly gathering at "\-",hich refreshments were to be served and to allow us to record the proceedings as samples of dialect speech. Two such gatherings were recorded, One in the living room of our local hosts, and the other in the home of an acquaintance. The fact that arrangements for the meeting were made by local people means that the groups were self-recruited. Participants in the first group included two sisters and a brother and their respective spouses. One of the men was a shopkeeper, one of the few in this category ,-vho claims local descent; his brothers-in-law were employed as craftsmen. All three men are quite literate compared to workmen elsewhere in the world and well read in public affairs. They are active in local politics and experienced in formal committee work. The second group included three craftsmen, friends and neighbors who worked in the same plant, and their wives. One of these had served as a sailor on a Norwegian merchant vessel for several years and spoke English. PartiCipants were all quite familiar with standard Norwegian, and our recorded conversations contain several passages where the standard was used in quoting nonlocal speech or in statements directed at us.

Methodologically, self-recruitment of groups is important for two reasons. It ensures that groups are defined by locally recognized relationships and enables the investigator to predict the norms relevant to their interaction. Furthermore, the fact that participants

SOCIAL MEANING IN LINGUISTIC STRUCTURE: CODE-SWITCHING IN NORWAY 225

have preexisting obligations toward each other means that, given the situation, they are likely to respond to such obligations in spite of the presence of strangers. Our tape recording and our visual observations give clear evidence that this in fact was what

occurred. Our strategy was to introduce discussion likely to mobilize obligations internal to

the group, thus engaging members in discussion among themselves. This proved to be relatively easy to do. When a point had been discussed for some time, we would attempt to change the subject by injecting new questions or comments. In doing this, we did not, of course, expect that our own interjections would predictably affect the speakers' choice of codes. Participants were always free to reinterpret our comments in any way they wished. Nevertheless, the greater the range of topics covered, the greater was the likelihood oflanguage shift.

As a rule, our comments were followed by a few introductory exchanges directed at us. These were marked by relatively slow sentence speeds, many hesitation pauses, and visual clues indicating that people were addreSSing us. Linguistically, we noted some s,vitching to the standard in such exchanges. After a brief period of this, if the topic was interesting, internal discussion began and arguments that referred to persons, places, and events we could not possibly be expected to have any knowledge about developed. The transition to internal discussion was marked by an increase in sentence speed and lack of hesitation pauses and similar clues. The tape recorder was run con- tinously during the gatherings, and after some time participants became quite oblivious to its presence.

Only those passages which were clearly recognizable as internal discussion were used in the analysis; all others were eliminated. The texts obtained in this way consist of stretches of free discussion on diverse topics. The following passages show that our hypothesis about the lack of connection between code switching and change in topic was confirmed.

Group I

Topic: Chitchat about local events

GUNNAR: ja de va ein sa kamm idaa~ein so. kamm me mcelka-sa sa hanj de va so. varmt inj pa rno i aar-ja, sa ea, de va no irre vent anjce dakk ma no ha meir enn di anjrann bestanjdi.

Yes, there was one who came today-one who came with milk-so he said it was so warm in Mo yesterday. Yes, I said, there is nothing else to be expected, you people must always have more than anybody else.

Topic: Industrial planning

ALF: her kunj ha vore eit par sann mellomstore bedriftce pa ein Jortifcemti manu so ha bescefiiace denna fllke detta sa ha aadd ledi amm I'injtcern.

There might have been here some medium-size plants employing forty to fifty men which then could offer work to those who have nothing to do in ,vinter.

226 JAN-PETER 8LOM AND JOHN J. GUMPERZ

Topic: Governmental affairs

OSCAR: vi jekk injJor denn Jorste injstilijin8~ ifra seikommitenn. "We supported the first proposal made by the Schei Committee."

Item 1 deals with a local topic in a somewhat humorous way; items 2 and 3 concern planning and formal governmental affairs. All these passages are clearly in the dialect. Of the phonological variables, [nil and [Ij] show the highest degree of palatalization and [a] and [re] the highest degree of retraction throughout. Morphophonemic dialect markers

(R) ." "0,, h'" " "dokk" " ." "h ''11 "'kk" "ijio are em one, sa W 0, lyfe not, a you, melT more, er ere, Je went, 1 fa "from." Even lexical borrowings from the standard such as injstiljing "proposal" and bedrift'" "plants" are clearly in dialect phonology and morphology. We find one Single instance of what seems to be a standard form: (B) melloml (R) imelja "middle." But this only occurs as part of the borrowed compound mellomstore "medium-size." In several hours of conversation with both groups, marked by many changes in topic, we have found a number of lexical borrOwings but not a clear instance of phonological or grammatical switching, in spite of the fact that all informants clearly know standard grammar.

While our hypothesis suggests that sWitching is constrained in those situations which allow only local relationships to be enacted, it also leads us to predict that whenever local and nonlocal relationships are relevant to the same situation, topical variation may elicit code SWitching. To test this, we selected members of a formerly quite active local peer group. For the last few years these individuals had all been at universities in Oslo, Bergen, and Trondheim. They returned home in the summer either for vacation or to take up local employment. In conventional interview sessions, all partiCipants claimed to be pure dialect speakers and professed local attitudes about dialect use. They thus regarded themselves as members of the local "team." As fellow students, however, they also shared statuses that are identified with pan-Norwegian values and associated with the standard. Our assumption then is that if topical stimuli are introduced which elicit these values, switching may result.

Three gatherings were arranged in the home of one of our informants. Refreshments were again served. Elicitation strategies were similar to those employed with the first two groups, and similar ranges of topics were covered. The examples cited here show that our hypothesis ,vas again confirmed.

Group II

Topic: Chitchat about drinking habits

BERIT: ja, ja, meEn I'i bj)'njt anjer l'eien dU-l'i bjynjt i barnelolen-sa I'i har de unjajort. Yes, yes, we started the other way, we started in the children's antialcoholic league.

So we have finished all that.

Topic: Industrial development

BERIT: jo da I'iss di bare fikk de te lann sen-sa e i l'eErif'all prisnil'oe heer i Rana skrudd honer enn de e vanlinl'is anner stann i lanne.

SOCIAL MEANING IN LINGUISTIC STRUCTURE: CODE-SWITCHING IN NORWAY 227

Yes, if they could only manage to make it profitable-so in any case the prices tend to be higher here in Rana than is common in other places in the country.

Topic: Informal statement about university regulations

OLA: mcenn no ha deemm lcempce pa de. But now they have relaxed that.

Topic: Authoritative statement about university regulations

OLA: de voel du mellom enJaemm saeks. You choose that from among five or six.

Comparison of Beries and Ola's first statement ,vith their second statements shows considerable shifting in each case. Thus Berie s second utterance has such unpalatalized forms as anner (vs. anjer), and raised and less retracted [a] in do. She also uses standard variables (B).fikk!(R)fekk, (B) viss!(R) vess, (B) vcertfall!(R) kvartjall, (B) heer! (R) her, etc. ala's second statement is characterized by (B) mellam!(R) imelja and (B) en!(R) ein. Similarly, his [tel in Jeemm and seeks is raised and fronted. In neither case is the shift to the standard complete-after all the situation never lost its informality. Berie s statement still contains dialect words like (R) lann! (B) lanne "to be profitable"; (R) stann! (B) steder "places"; and Ola has (R) vcell (B) velaer "to choose." What we see then is a breakdown of co-occurrence rules, an erosion of the linguistic boundary between Ranamal and Bokmal. The tendency is to switch toward standard phonology while preserving some morpho- phonemiC and lexical dialect features of (R). Features retained in this manner are largely those which also occur in other local dialects and to some extent also in Nynorsk. They have thus gained some acceptance as proper dialect forms. Those characteristics which locals refer to as broad speech, i.e., those that are known as local peculiarities, tend to be eliminated.

It must be noted as well that Berit and ala also differ in their pronunciation of the phonological variables. Ola's normal pronunciation shows the strong palatalization of consonant and extreme vowel retraction characteristic of most residents. Berie s normal pronunciation has medium palatalization and medium retraction. Both, however, switch in the same direction, in response to similar situational and topical clues, and this agreement on the rules of stylistic manipulation is clearly more important in this case than the mere articulatory difference in Beries and Ola's speech.

The social character of the style switch was clearly revealed when the tape-recorded conversations were played back to other Hemnes residents. One person who had been working with us as a linguistic informant at first refused to believe that the conversations were recorded locally. When he recognized the voices of the participants, he showed clear signs of disapproval. Apparently, he viewed the violation of co-occurrence rules as a sign of what is derogatorily called knot "artificial speech" in colloquial Norwegian. Some of the participants showed similar reactions when they heard themselves on tape. They promised to refrain from switching during future discussion sessions. Our analYSis of these later sessions, however, revealed that when an argument required that the speaker

228 JAN-PETER BLOM AND JOHN J. GUMPERZ

validate his status as an intellectual, he would again tend to use standard forms in the manner shown by Berit and Ola. Code selection rules thus seem to be akin to grammat- ical rules. Both operate below the level of consciousness and may be independent of the speaker's overt intentions.

Additional information about usage patterns in group III was provided through a fortunate accident. One of Our sessions with this group was interrupted by a [ ... ] [local youth with a cognitive disability], who has the habit of appearing in people's homes to solicit assistance for his various schemes. Here are some examples of remarks addressed to him by Berit and Solveig, of all the members of the group the most prone to use standard forms. Her normal pronunciation shows the least amount of consonant palatalization. She is socially more marginal to Hemnes than other members of the group.

Group III

Topic: Talking to a (. . .] local youth

BERIT: e de du sa vikarier Jorr han) no. Are you a stand-in for him now?

SOLVEIG: hanj kanj jo jett glite, haj kanj no va me. He is good at word games, he should participate.

Both Berit and Solveig's pronunication in these examples becomes identical with the ordinary speech of Ola and of the members of group I. The extreme palatalization of [nj] and the lowering of [a] is not normal for them; they clearly are talking down in this case. Their stylistic range, as "veIl as their facility in switching, seem to be greater than those of the others.

In comparing the behavior of the first two groups with that of group III, we find two different kinds of language-usage patterns. All three groups speak both the dialect and the standard. Groups I and II, however, show only situational switching. When members talk to each other, differences of formality or informality to topic are reflected only in the lexicon. Pronunciation and morphology do not change. Those groups shift to (B) phonology and grammar only when remarks are addressed directly to us who count as outsiders or in indirect quotes of such matters as government rules, on officials' state- ments, etc. In such instances of situation switching, therefore, Ranam~n and Bokmal are kept separate throughout by strict co-occurrence restrictions. In group III, however, deviation from the dialect results both from metaphorical and situation switching. Metaphorical switching, furthermore, involves a breakdown of the co-occurrence restrictions characteristic of situational shifts.

The dialect usage of locals, on the one hand, corresponds to their view that the two varieties are distinct, and to their insistence on maintaining the strict separation of local and nonlocal values. For the students, on the other hand, the distinction between dialect and standard is not so sharp. Although they display the same general attitudes about the dialect as the team of locals, their behavior shows a range of variation rather than an alternation between distinct systems. It reflects a de facto recognition of their own nonlocal identification.

SOCIAL MEANING IN LINGUISTIC STRUCTURE: CODE-SWITCHING IN NORWAY 229

A fourth conversational group further illustrates the internal speech diversity in the community. The principal speakers here are two men, A and B, and C, who is A's wife. All come from families who tend to dissociate themselves from the egalitarian value system of the local team. Their normal style of speech was BokmiU for remarks directed at us as well as for in-group speech. Only in a few instances when A began telling local anecdotes did he lapse into Ranamal. (R) forms were introduced as metaphorical switches into what \vere basically (B) utterances to provide local color, indicate humor, etc., in somewhat the same way that speakers in group III had used (B) forms in (R) utterances. .

In the course of the evening A and C's teen-age daughter joined the conversation. She expressed attitudes toward the dialect which are quite similar to those of the students in group III and thus are somewhat different from those of her parents. The few samples we have of her speech show (R) phonology similar to that of Berit and Solveig in group III.

Although the picture of language usage derived from the four groups seems at first highly complex, it becomes less so when viewed in relation to speakers' attitudes, interactional norms, and local values. All Hemnes residents have the same repertoire. Their linguistic competence includes control of both (R) and (B) rules. They vary in the way in which they use these rules. Expressed attitudes toward (R) and (B) do not provide an explanation for these differences in speech behavior. The most reasonable explanation of the ways in which these groups differ seems to be that the dual system of local values, differences in individual background, and the various social situations in which members find themselves, operate to affect their interpretation of the social meaning of the variables they employ.

[ ... J

REFERENCES

Barth, Fredrik (1966) Models if Social Oraonization. Royal Anthropological Institute of Great Britain and Ireland. Occasional Papers. London.

Christiansen, Hallfried (1962) Malet i Rona. Oslo: Institut for Sociologi, Universitetet i Oslo. Friedrich, Paul (1972) Social context and semantic feature: the Russian pronominal usage. In John

Gumperz and Dell Hymes (eds) Directions in Sociolinauistics: The Etlmoaraph)' if Communication. New York: Holt, Rinehart and Winston, 270-300.

Coffman, Erving (1964) The neglected situation. In John J Cumper;; and Dell Hymes (cds) The Ethnoaraph)' if Communication. American Anthropoloaist 66, 6, pt. II: 133-137.

Gumperz, John J. (1964) Linguistic and social interaction in two communities. In John J Gumperz and Dell Hymes (cds) The Ethnograph), if Communication. American Anthropologist 66, 6, pt. II: 137-154.

Labov, William (1966) The Social Stratification if Enalish in New York Ci9'. Center for Applied Linguistics, Washington, DC.

Leach, Edmund (1954) Political Systems if Highland Burma. Cambridge, Mass.: Harvard University Press.

<~

~

230 JAN·PETER BLOM AND JOHN J. GUMPERZ

QUESTIONS

L 2<

3.

4. 5. 6. 7. 8. 9.

What is the linguistic situation in Hemnesberget?

What evidence do Blom and Gumperz provide in support of their view of dialed features as variables within a single grammatical system?

What factors influence selection patterns of the two varieties and how did these patterns

come about?

What does the term social meaning refer to and how does it relate to Ranamal and Bokmal? What is the social structure like in Hemnesberget and how does it relate to language use? What contextual constraints influence language use?

What are situational and metaphorical code-switching?

What type of code-switching occurs in groups I and II of the experiment?

What type of code-switching occurs in group III of the experiment?

Do you often hear people code-switch? What languages are used? What social meaning do

these languages and code-switching activities have to you? What social meaning might they

have to the conversationalists themselves?