Accommodation Theory

by Dimitrios Thanasoulas, B.A.

Each one of us is aware that our style of speech changes in the twinkling of an eye, as it were, depending on a wide range of variables such as the setting, the topic of discourse, the person we are interacting with, the purpose of the interaction, and so on. For instance, we tend to speak more slowly when conversing with foreigners, or use grammatically simple language with babies or children (babytalk). In other words, we accommodateto others by adjusting our communicational behaviour to the requisite roles that participants are assigned in a given context. In the present study, we will content ourselves with accommodation theory or "accommodative processes" (Giles & Coupland, 1991) in relation to identity, with a view to shedding light on the different ways in which speakers may manipulate language "to maintain integrity, distance or identity" (ibid, p. 66). Furthermore, we will address ourselves to the reasons why low-prestige language varieties and stigmatised accents persist in a society where the use of standard speech confers prestige and power on its users. In short, the study of accommodation theory may, on the one hand, reveal the extent to which language impinges on our lives, resulting in the maintenance or breakdown of human relationships, and on the other give useful insights into the tendency for different varieties to evoke or "trigger" different perceptions of their speakers.

That language is "socially diagnostic" (Giles & Coupland, 1991) is manifest in everyday conversations. How many times have we come across a "different" accent or pronunciation of an individual sound without adopting a critical stance or making guesses as to the speaker's non-linguistic characteristics, such as status, education, class, or even intelligence? In fact, the slightest nuance in pronunciation, not to mention stylistic discrepancies, can as often as not "have evaluative repercussions for its utterer" (ibid, p. 32). Huspek (1986: 158, cited in Giles & Coupland, 1991: 32) contends that, if someone says, "I went joggin' this morning" instead of "I went jogging this morning," chances are that in the first case he will be perceived as being of lower rank than in the second case. As Giles & Clair (1979: 17) note, "language is not a homogeneous, static system. It is multi-channelled, multi-variable and capable of vast modifications from context to context by the speaker, slight differences of which are often detected by listeners and afforded social significance." Given the fact that even the most trivial aspects of speech and pronunciation can take on crucial importance, it stands to reason that individuals, consciously or unconsciously, should, among other things, seek or eschew identification with others through language. It is in this light that accommodation theory has become an important, albeit controversial and besetting, issue in sociolinguistics and social psychology.

It is a truism that accommodating to others' speech may prove beneficial or detrimental, in the long run. For example, immigrants whose command of standard English or any other language is not "up to scratch" is bound to suffer discrimination and prejudice on the part of teachers and society at large, which puts paid to their educational and career prospects. Moreover, adapting our speech patterns (pronunciation, speech rate, content etc.) to those of our interlocutors can exert a tremendous influence on our career prospects and prestige, or even affect the judicial outcome of a trial. At any rate,

…accommodation is to be seen as a multiply-organized and contextually complex set of alternatives, regularly available to communicators in face-to-face talk. It can function to index and achieve solidarity with or dissociation from a conversational partner, reciprocally and dynamically (Giles & Coupland, 1991: 60-61).

Accommodation theory or "interpersonal accommodation theory" has sprung from the awareness that speakers are not merely "incumbents" (Runciman, 1998) of roles imposed on them by society but rather as inquirers attempting to comprehend themselves and others. Viewing individuals as objects called upon to modify their speech in accordance with socially prescribed norms leaves much to be desired, in so far as it ignores the interactants" feelings and motives, which undoubtedly inform and permeate the production and interpretation of their verbal output. Accommodation theory focuses on the interactive aspects of communication and highlights its "negotiative" nature. So, in order to do justice to this model of speech diversity—that is, interpersonal accommodation theory—we should examine four social psychological theories that actually constitute it and account for people's tendency to converge towards or diverge away from the speech of others: similarity-attraction, social exchange, causal attribution, and Tajfel's theory of intergroup distinctiveness.

A very common modification of speech is what has been dubbed as convergence. This term refers to the processes whereby two or more individuals alter or shift their speech to resemble that of those they are interacting with. By the same token, divergence refers to the ways in which speakers accentuate their verbal and non-verbal differences in order to distinguish themselves from others. As was mentioned above, there is a tendency for people to become more alike in terms of linguistic, prosodic or non-verbal features, including pronunciation, utterance length, pauses, speech rates, vocal intensities, as well as facial expressions and the "intimacy of their self-disclosures" (McAllister and Keisler, 1975, cited in Giles & Clair, 1979: 46). What is more, many studies have emerged showing that convergence in interethnic settings exists at six years of age (Aboud, 1976, cited in Giles & Coupland, 1991), though a deeper knowledge of the sociocultural norms pertaining to language use is acquired throughout childhood and adolescence.

Therefore, we could contend that the "universality" of convergence strategies may well point to people's perennial need for social approval and mutual intelligibility, which is the underlying assumption proposed by similarity-attraction theory. "The more similar our attitudes and beliefs are to certain others", so the theory goes, "the more likely it is we will be attracted to them" (Byrne, 1969, cited in Giles & Clair, 1979: 47). For example, when an acquaintance of ours pronounces "leak" as "lick," it is probably the case that, out of politeness and because of the need to maintain solidarity and rapport, we will ignore her mistake and go on with the conversation. The same applies to affective language and phatic expressions such as Hello, how are you this morning? or I'm awfully sorry, where two or more persons converge towards one another both verbally and non-verbally (e.g., by mutual gazing or smiling etc.), in order to signal that they are on the same wavelength and wish to maintain good relationships. Interestingly, as Natalé (1975, cited in Giles & Clair, 1979) believes, individuals with high needs for social approval and intelligibility have the propensity for converging more to others" pause length and vocal intensities than those with higher self-esteem and lower needs for approval.

Moving away from the possible rewards attending an act of convergence, such as an increase in social approval, we must also consider the costs involved, i.e., an increased effort to appear likeable and friendly, and the concomitant loss of personal integrity and identity that such an effort may entail. Of course, social exchange theory suggests that speakers and listeners share "a common set of interpretative procedures which allow the speaker's intentions to be (i) encoded by the speaker, and (ii) correctly interpreted by the listener" (Giles & Clair, 1979: 46-47). Besides, it presupposes that, "prior to acting, we attempt to assess the rewards and costs of alternate courses of action" (Homans, 1961, cited in Giles & Clair, 1979: 48). Thus, engaging in speech convergence may incur more rewards than costs. For instance, in England "Received-Pronunciation" (RP) speakers are looked upon as more intelligent, serious and self-confident than regional accented speakers. In the same vein, there is empirical evidence that people react more favourably to those converging towards them, while it is almost always the case that the very same persons judged favourably in the first case will be denigrated as uneducated, uncouth and socially incompetent when using vernacular varieties, as the matched-guise technique developed by Wallace Lambert and his associates has shown (Fasold, 1987: 149-150). Moreover, the act of convergence, upward and downward, may stand one in good stead. Consider the case where a young employee, aspiring to a salary rise or promotion, may converge upwards towards his boss by using formal language; or when an employer converges downwards towards his workers in order to win their approval.

Nevertheless, observing people's behaviour and taking it at face value is not what interpersonal communication is all about. Causal attribution theory proposes that, when we interact with others, we engage in an interpretative process, evaluating the individuals in terms of the possible motives that we attribute as the cause of their action. For example, we do not just observe an affluent man helping the poor and instantly become enraptured by his kindness and generosity. Rather, we tend to consider his motives first. In this light, if we attribute to him a personal gain from this act, then we may take a dim view of his behaviour, judging him negatively as a shallow and Machiavellian opportunist. By the same token, speech convergence may not be favourably received when attributions of speakers" intentions are negative. For instance, an experiment has shown that, when French Canadian listeners attributed an English Canadian's shift to French to his desire to achieve solidarity, they judged him favourably. However, when his act was attributed to pressures forcing him to converge, less positive feelings were evoked (Simard, Taylor and Giles, 1976, cited in Giles & Clair, 1979: 50).

Within the context of the theory of intergroup distinctiveness, Tajfel proposes that when different groups come in contact, there is a tendency for them to compare themselves on the grounds of abilities, possessions, personal traits, accomplishments, and so forth. According to his theory, these "intergroup social comparisons" will assist individuals in forging their group image and positive ingroup distinctiveness. It may be the case that individuals seek solace in the knowledge that they are part of groups which enjoy some primacy and prestige. Given that speech is to be seen "as an identity adjustment made to increase group status and favourability" (Edwards, 1985: 152), we could argue that in situations when group membership has to be accentuated and supported, speech divergence may be an important strategy for distinguishing oneself from members of other groups. An example of speech divergence is given in Holmes (1992: 257):

A number of people who were learning Welsh were asked to help with a survey. In their separate booths in the language laboratory, they were asked a number of questions by an RP-sounding English speaker. At one point this speaker arrogantly challenged the learners' reasons for trying to acquire Welsh which he called a "dying language which had a dismal future". In responding to this statement the learners generally broadened their Welsh accents. Some introduced Welsh words into their answers, while others used an aggressive tone. One woman did not reply for a while, and then she was heard conjugating Welsh verbs very gently into the microphone.

As is flagrantly obvious, the respondents diverged from the speech style and language of the person addressing them because they felt threatened and denigrated. "They disagreed with his sentiments and had no desire to accommodate to his speech" (ibid, p. 257).

That "non-convergence" or divergence may act as a symbol whereby members of an ethnic group can signal their intention of maintaining their distinctiveness is further exemplified by the decision of the Arab nations to issue an oil communiqué to the world in Arabic, thus making a political statement. They no longer wished to accommodate to the Western English-speaking powers. The same applies to Maori dissidents who, despite the fact that they can speak English fluently, they have insisted on speaking Maori in court.

As a matter of fact, there are cases where divergence between groups is expected, its absence being construed as a token of dissociation from intergroup values and norms and thus as a signal of unwarranted friendliness and allegiance to the opposing group.

In certain intergroup encounters, members of opposing or competing groups may expect nothing less than some linguistic divergence (rather than convergence) if only to remind themselves that they indeed have a dispute or difference (cf. Doise, Sinclair and Bourhis, 1976). Failure to confirm these expectations might indicate that something was "wrong", that the intergroup situation had changed somehow without one party being aware of it… Indeed, one can imagine situations in which the "right" amount of divergence might elicit co-operation where convergence would not (Giles & Clair, 1979: 63).

Therefore, divergence can be a tactic of intergroup distinctiveness at the disposal of people seeking a positive social identity. On an interpersonal note, overdoing divergence—as well as convergence—may offend others. Scotton (1985, cited in Giles & Coupland, 1991) introduced the term "dis-accommodation" to refer to the shift of registers by certain people in repeating something uttered by their interlocutors. For example, a thirty-year old man might say, "OK, mate, let's get it together at the bar at 6:00 tomorrow", and receive the reply from a fifty-year old man, "Fine, young man, we'll meet again, 18:00, at your house tomorrow." Maintaining one's idiosyncratic speech patterns may be spontaneous and inherently unexceptionable, but when it comes to communication, one may be frowned upon as disdainful, pompous and unapproachable when systematically diverging away from others' speech. Let us give an example of over-convergence. Imagine a situation where a person converges towards the pronunciation of someone talking in a lisp. It is highly unlikely that she will be regarded as polite or as signalling that she is on the same wavelength, seeking to achieve solidarity and good rapport with her interlocutor. Rather than sounding considerate and friendly, she will be perceived as patronising or even unctuous.

Having dilated upon these four social psychological theories, i.e., similarity-attraction, social exchange, causal attribution, and Tajfel's theory of intergroup distinctiveness, we have moved "closer to the interactive interface between speaker and listener, and [arrived] at the conception of the procedures employed by them in the production and interpretation of speech style shifts" (Giles & Clair, 1979: 53). Yet, we have to consider some cases when the speaker may be perceived as optimally accommodating. Giles & Smith (1979, cited in Giles & Coupland, 1991) presented eight versions of a tape recording to several English subjects. The taped voice was that of a Canadian, who either converged or diverged on three dimensions, i.e., pronunciation, speech rate, and speech content. It was shown that listeners appreciated convergence on each of the aforementioned levels separately, while they dismissed full convergence on all three levels as patronising. The bottom line is that when the speaker converged in terms of content plus speech rate, he was most favourably evaluated.

Besides, Giles & Smith argued that there may also be optimal rates of convergence and divergence. More specifically, Aronson and Linder (1965, cited in Giles & Coupland, 1991) proposed "gain-loss" theory of attraction, according to which people feel stronger liking for those whose respect they are acquiring than for those whose respect they already enjoy. What can be extrapolated from this is that convergence is preferable and more effective when taking place incrementally than all at once. "Gain-loss" theory also claims that people dislike those whose respect they have lost rather than those who have never held them in high regard. Looked at from an "accommodation theory" perspective, individuals are apt to disapprove of those who diverge sequentially away more than those who diverge all at once.

We have hitherto been concerned with two basic accommodation strategies—convergence and divergence—which are deployed by individuals to signal identification with, or dissociation from, the communication patterns of others. In this light, we could say that these strategies are the linguistic realisations of deeper goals and orientations that individuals tacitly negotiate. Thakerar et al. (1982, cited in Giles & Coupland, 1991) have made the distinction of psychological versus linguistic accommodation, defining the former as individuals' integrative or dissociative orientations to others, and the latter as the speech strategies realising these orientations.

Apart from this, Thakerar et al. (1982, cited in Giles & Coupland, 1991) have suggested that convergence and divergence are not only affective phenomena but may also function as cognitive organisation devices.

The cognitive organization function involves communicative features being used by communicators to organize events into meaningful social categories, thereby allowing the complex social situation to be reduced to manageable proportions. In this way, speakers may organize their outputs to take into account the requirements of their listeners; listeners may select from this discourse and organize it according to the cognitive structures most easily available for comprehension (Brown and Dell, 1987, cited in Giles & Coupland, 1991).

Clear examples of such devices are "babytalk"—which fulfils the cognitive organisation function of simplifying one's output—and a sociologist's attempt to make himself understood to people who are not versed in the jargon.

A question germane to the present study is, "why do low-prestige language varieties persist?". Why is it that certain groups of people insist on using vernacular varieties, even though, in doing so, they may run into intractable difficulties, in terms of educational and career prospects, prestige and status, and so forth? In fact, one might expect these varieties to disappear, given that the high prestige standard is used predominantly by the social groups with the highest status. Yet, rather than deteriorating, vernacular dialects may - and in some cases have - become a regional standard over a high status variety, an oft-quoted example being Greece, where katharevousa, the standard variety, has been replaced by dhimotiki, a vernacular.

Attempting an answer to the question posed, Gubuglo (1973, cited in Giles & Clair, 1979) ascribes the preservation of non-standard dialects to the value of language as a symbol of group identity. Furthermore, Taylor and associates (Taylor, Bassili and Aboud, 1973; Taylor, Simard and Aboud, 1972, cited in Giles & Clair, 1979: 147) believe that language is a critical dimension of identity, while Patterson (1975, cited in Giles & Clair, 1979) holds that, just as ethnicity is a matter of choice, accent or dialect adoption is a matter of conscious choice.

Thus, although regional, ethnic, and lower-class individuals have limited access to opportunities for acquiring the prestige variety compared to members of the high status groups, much of the failure of these individuals to profit from whatever opportunities are available is due to counter-acting pressures favouring their native speech styles (ibid, p. 148).

Indeed, the more important it is for a particular group to maintain its cultural distinctiveness, the more salient language becomes. In a way, "language functions as a very sensitive filter through which one's perception of self, own group, and others must pass" (ibid, p. 187).

Nevertheless, depending on individuals' motivation and purposes, such "low-prestige" varieties may lose ground to the standard variety, if the speakers of non-standard forms choose to move away from the contexts in which these are vernaculars. In other cases, many speakers of regional dialects may become bi-dialectal, shifting their speech according to the situations they are in.

To sum up, we could say that accommodation theory has helped us understand why individuals speak the way they do, accounting for the manner in which they interpret their own roles and those of their interlocutors, as well as the procedures they resort to in order to act meaningfully. What is more, the "accommodative processes" that people employ may fulfil the function of attenuating or accentuating their social identity, which inescapably opens up new vistas of study.


  • Edwards, J. 1985. Language, society and identity. Oxford: Blackwell.
  • Fasold, R. W. 1987. The Sociolinguistics of Society. Oxford: Blackwell.
  • Giles, H. & Clair, R. 1979. Language and Social Psychology. Oxford: Blackwell.
  • Giles, H. & Coupland, N. 1991. Language: Contexts and Consequences. Keynes: Open University Press.
  • Holmes, J. 1992. An Introduction to Sociolinguistics. London: Longman.
  • Runciman, W. G. 1998. The Social Animal. Great Britain: Harper Collins Publishers. ESL Reviews & Articles© Dimitrios Thanasoulas 1999