«Journal of Sociolinguistics 4/3, 2000: 323±347 Styling the worker: Gender and the commodi®cation of language in the globalized service economy1 ...»
Styling is used ± either on its own or in combination with scripting ± because specifying a standard form of words does not on its own ensure the kind and degree of standardization many service organizations, including many call centres, are trying to achieve. Scripting standardizes what is said, but styling is an attempt to standardize how it is said, addressing the many aspects of # Blackwell Publishers Ltd. 2000 332 CAMERON spoken interaction that are not readily represented in a written script. Consider, for example, the following remark made by a supervisor at an airline reservations centre to the sociologist Steve Taylor (she is discussing the use of taped
calls for the purpose of `counselling' operators):
A lot of the time it isn't what they say, it's the tone in which they say it... I will play something and I'll just stop it and go, `shall we listen to that again?', rewind it and then they'll go, `I didn't know I said it like that'. It makes them analyse themselves and really wake up to their mistakes (Taylor 1998: 93, emphasis in original).
The object of concern here ± tone of voice ± lies beyond the reach of scripting, but it is nevertheless assumed by the supervisor to be susceptible to judgements of correctness (cf. her use of the word mistakes). Both scripting and styling are intended to ensure that workers speak `correctly' from the company's point of view; one takes over where the other leaves o.
From a linguist's point of view the concerns embodied in styling rules fall into two main categories. Firstly, as in the above example, attention is given to the operator's use of her/his voice, with a particular focus on suprasegmental phenomena such as voice quality and intonation. Secondly, emphasis is placed on various aspects of the management of interactive spoken discourse. Operators may be reminded for instance to avoid gap and overlap in turn transitions, to use minimal responses frequently, to ask `open' questions and to pause so that callers can assimilate important information.
Some of these considerations (e.g. the phrasing of questions) can be incorporated into a pre-written script, but many cannot (e.g. the placement of minimal responses and turn transitions, which depends on the behaviour of the caller).
Probably the most important instruments of styling are the checklists used in many centres for purposes of assessment by supervisors, managers and `mystery callers' (that is, outsiders employed to perform `spot checks' by posing as real callers and then logging their assessment of the operator's performance). Here, for example, is a selection of the contents of a 12±point checklist used in the assessment of operators at the credit authorization centre in my sample (a
centre, incidentally, which also scripts call routines exhaustively):
. Smiling. Does the member of sta answer the phone with a smile?
. Pitch. The depth of pitch in the sta 's voice will determine the degree of sincerity and con®dence associated with the message that they are giving the caller.
. Volume. Ensure sta are not shouting or hardly audible.
. Pace. Ensure the member of sta is not dragging out the sentences nor speeding through it [sic].
. Acknowledge. Sta can let the caller know they have understood them by making simple acknowledgement sounds, if the caller is not acknowledged in this way they will presume they have not been understood and repeat themselves.
understands what `depth of pitch' refers to), but it does at least specify which aspects of performance operators and assessors are expected to be attentive to.
(Lists like this one are typically used in both formal appraisal and more informal regular coaching of individual operators by their supervisors. They also inform the preliminary training of operators.) Having established what I mean by `styling' and how it is embedded in the call centre regime, I now turn to a more detailed examination of the linguistic characteristics of the preferred style. I will seek to show that the style is gendered, produced through a consistent and deliberate preference for ways of speaking that are symbolically coded as `feminine' (and that in some cases are also empirically associated with women speakers).
CALL CENTRE STYLE AND `WOMEN'S LANGUAGE'As is well known, 25 years ago Robin Lako (1975) elaborated a notion of `women's language' (WL), a register or, in Eckert's sense, a `style' characterized by linguistic features such as the use of `weak' expletives and lexical items like charming, divine, rising intonation on declaratives, tag-questions in contexts where the speaker is not checking information, etc. Subsequent empirical investigations of the `Lako hypothesis' produced a copious literature, the import of which is perhaps most succinctly summarized by saying that not all women use WL and not all WL-users are women. This however did not deter scholars from advancing alternative proposals about women's style of speaking and how it diers, on average, from men's. For example, one general claim widely canvassed in the 1980s and 1990s was that women are more cooperative conversationalists and more sensitive to the face-wants of others (Coates 1996; Holmes 1995; Tannen 1990). This dierence has been invoked to explain women's use of an array of discourse features such as supportive simultaneous speech, precision-timed minimal responses and questions whose function is to show interest in or engage the participation of others, hedging and indirectness used to mitigate face-threat, and so on.
This brief excursion into the history of language and gender studies is relevant here, because it is evident that the products of the research tradition inaugurated by Lako have ®ltered steadily, though selectively, into popular consciousness. This process has produced a lay notion of `women's language' that is an amalgam of long-established folk-beliefs, elements of the early Lako hypothesis, popularized accounts of more recent ®ndings, and new, or at least reworked, stereotypes disseminated via popular psychology and self-help texts.
However inaccurate it may be as an empirical description of the way women `really' speak, and however unsatisfactory it may appear from the perspective of academic scholarship, this notion of `women's language' provides a powerful symbolic `meaning resource' for `stylistic agents' to draw on. In the following discussion I will seek to show in more detail how various elements of the symbolic construct `women's language' are appropriated and recombined in the # Blackwell Publishers Ltd. 2000 334 CAMERON call centre context to produce a particular service style. The discussion is based on materials (e.g. training manuals and appraisal criteria) I collected from four call centres in my sample, and it focuses on concerns that recur across those materials.
One concern that is highlighted in all the materials I collected is with the styling of the operator's voice. Two instructions on vocal performance are invariably given: that operators should smile ± even though, obviously, they are invisible to their interlocutors ± and that they should use an `expressive' intonation. What the instruction to smile actually means is that the routine (or sometimes just part of it, e.g. the opening) should be performed with the lips in a smile posture. `Expressive' intonation means emotionally expressive, and is explicitly contrasted to intonation which will be heard as monotonous or uninvolved.
`Expressive' intonation projecting attitudes/emotional states Our commitment is to give the caller an impression of excitement, friendliness, helpfulness and courtesy. Your telephone manner should sound as if you have been waiting for that particular call all day. You must never sound bored on a call (directory assistance centre employee manual) The objective at the beginning of a call is to demonstrate sincerity and warmth. Try to make the caller feel you are there for them... [avoid] a disinterested, monotonous tone to voice (performance guidelines, auto insurance centre) It has been argued that both smiling and using expressive intonation are symbolically feminine behaviours. In the case of smiling, nonverbal communication researchers point out that it is not simply a spontaneous expression of pleasure but often functions, especially with non-intimates, to signal deference or appeasement. In the words of Nancy Henley, the smile is `understood as a gesture oered upward in the status hierarchy' (1986: 171). This analysis of what smiling means has in turn been linked with ®ndings suggesting that women smile more than men and that they are more likely to return smiles than men (Henley 1986: 175±178). It has also been linked with the observation that women are routinely expected to smile, and sometimes publicly castigated by complete strangers if they do not smile. Shulamith Firestone (1970: 90) once proposed a `smile boycott' as a form of feminist political action;
# Blackwell Publishers Ltd. 2000 STYLING THE WORKER 335 in 1999 female ¯ight attendants employed by Cathay Paci®c airlines threatened to take industrial action in a dispute on pay and conditions by refusing to smile at passengers for one hour of every ¯ight. Such actions are meaningful precisely because of the existence of strong symbolic links between smiling, femininity and subordinate status.8 As for expressive intonation, it is both a stereotype and in some cases an empirical ®nding that female speakers exploit a broader pitch range, in other words tend less to monotony. This characteristic has been used in the past to label women as over-emotional and lacking in authority, tempting women like Margaret Thatcher to deliberately reduce the pitch range they use. The fact that vocal expressiveness is valued in service-work might suggest that authority is not among the qualities workers are expected to display.
If we consider the sorts of emotional or attitudinal states operators are instructed to project through their intonation, we see references in the above examples to warmth, sincerity, excitement, friendliness, helpfulness, con®dence.
These are not inherently gendered qualities, but overall they produce a style of service which is strongly aective ± that is, not just neutrally polite and ecient, but based on the expression of positive feelings towards the customer.
Again, it has been argued that overt displays of positive aect, or of any emotion other than anger, are culturally coded as `feminine' rather than `masculine' (Gervasio and Crawford 1989).
Other recurrent styling concerns are to do with the management of interpersonal relationships through strategic choices at the level of discourse.
One common instruction, for example, is to create rapport with callers, while another is to display empathy with them. In this example these (related)
concerns are combined in the following, quite lengthy recommendation:
Rapport/empathy Creating a rapport and showing empathy is about adding the human touch to a business call relationship... This means treating the caller as a person, recognising their situation and building a genuine conversation to re¯ect this.... Use language which conveys understanding of and empathy for the caller's individual situation, e.g.
`are you OK?' `was anyone hurt?' `that must have been very distressing for you' (performance guidelines, auto insurance centre) Here, two main discourse strategies are suggested. One is asking questions to show concern for the caller and encourage her/him to air her/his feelings about the incident that prompted the call (in this context, a trac accident). The other is the technique known to communication trainers as `mirroring', which means trying to demonstrate awareness of the interlocutor's mood and re¯ect it back to her/him in your own verbal and nonverbal behaviour. It is, of course, a common stereotype that women are better than men at inferring others' feelings from their outward behaviour, which is a precondition for successfully displaying empathy. The association of rapport-building with women's talk appears in # Blackwell Publishers Ltd. 2000 336 CAMERON many sources, notably Deborah Tannen's (1990) aphorism that men do `report talk' and women do `rapport talk'.
Another issue that is often addressed in call centre styling materials is the use of minimal responses. Concern about this aspect of interaction might seem to be motivated primarily by the need to make operators aware of speci®c constraints aecting telephone talk ± that is, since there are no visual cues, verbal backchannelling is necessary to reassure the caller that the operator is still present and listening actively. However, the following example shows that the writer realizes there is more to the use of minimal responses than simply keeping the channel of communication open.
Minimal responses Use words of acknowledgement: yes, OK, thank you, I understand, I see.... [avoid] disruptive, disinterested or challenging use of listening acknowledgements, and using the same listening acknowledgement throughout the call (performance guidelines, auto insurance centre) This is a recommendation to use minimal responses supportively: they should not be inserted where they will disrupt interaction, connote lack of interest or disagreement. It may be recalled here that some researchers (Fishman 1983;
Reid 1995) have found women not only using more minimal responses than men, but also timing them more precisely, to coincide with or immediately follow the completion of the point they are responding to. The use of delayed minimal responses, which may suggest inattention, lack of interest or disagreement, has been associated more with male speakers. Once again, what is being recommended here would seem to be gendered, matching what is believed and what in some cases has been found to be women's rather than men's behaviour.