ChapterPDF Available

Active listening in conversation: gender and the use of verbal feedback

Authors:
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
1
1 Introduction
Conversation consists of much more than alternating turns at talk: rather, it is a collaborative,
endeavour which requires participants to engage actively as both ‘speakers’ and ‘listeners’ in the
sequential construction of an interaction. Listenership therefore involves much more than simply
being a passive recipient of what the current speaker is saying: it means taking an active part in
the ongoing construction of the interaction. This is empirically observable in the way listeners
use a variety of verbal responses and non-verbal cues, as well as in the pragmatic use of silence
and pauses, all of which can vary in frequency, distribution and intensity to fulfil a complex array
of interactive, epistemic and social functions.
Supportive verbal feedback takes many forms, ranging from brief vocalisations such as
laughter, minimal responses such as mhmm and yeah and other brief expressions of overt or
implicit support or agreement, through to longer sequences of various sorts, including
backchannel utterances and cooperative overlaps. Verbal feedback occurs regularly and
frequently in all types of conversational interaction, closely aligned with non-verbal signals
such as gestures, head nods, facial expression and gaze, and is a primary resource for
engaging in active listenership. At a minimum these devices serve to signal the listener’s
alignment with the current interactional activity, but they may also help to construct more
substantive forms of involvement and affiliation between interlocutors.
Using tools from conversation analysis, interactional sociolinguistics and pragmatics, analysts
of spoken discourse have developed an increasingly sophisticated understanding of how
verbal listener responses function in their sequential context (e.g. Bublitz 1988; Coates 1989;
Farr 2003; Gardner 1997, 1998, 2001, 2007; Jefferson 1984, 1993, 2002; McCarthy 2003;
Norrick 2011; Orestrom 1983; Schegloff 1982; Schffrin 1987; Stenstrom 1987, 1994; Stivers
2004, 2008; Stubbe 1994, 1998; Ward & Tsukahara 2000, 2003; Xudong 2008). More
recently, with the increasing availability of video corpora and analytic tools, the focus of
interest has widened further to encompass micro-level multimodal analysis of active
listenership, specifically the fine articulation between verbal responses and non-verbal cues
(e.g. Carter & Adolphs 2008; Knight & Adolphs 2008; Stivers 2008).
The focus of this chapter is on gendered patterns of variation in the use of supportive verbal
feedback in a sample of conversational data from the Wellington Corpus of Spoken New
Zealand English (WSC).1The original analysis of this data was carried out between 1995and
1998 as part of a wider-ranging study of pragmatic variation in New Zealand English
discourse. An analytic framework for the classification of supportive verbal feedback in
corpus data was developed by the author as part of this project (see section 3), and was
subsequently applied to a quantitative and qualitative analysis of ethnic and gender variation
in a matched sub-sample of the WSC (see Stubbe 1996; Holmes & Stubbe 1997; Stubbe
1998).2
1 The Wellington Corpus of Spoken New Zealand English consists of one million words of spoken New Zealand
English made up of speech excerpts of 2000-2500 words. All contributors are speakers of New Zealand English,
from a representative range of gender, age, and social and ethnic backgrounds (Holmes, Vine & Johnson 1998).
See Sigley (Chapter X , this volume) for an overview of the text-category markup of this corpus.
2 This chapter provides a more detailed account of the gender component of this previous analysis than has been
published previously, including some previously unpublished results, and draws on relevant material in later
publications by the author and colleagues (e.g. Stubbe 1999, 2003; Stubbe, Holmes, Vine & Marra 2001; Stubbe,
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
2
A brief overview of the relevant literature on gender and verbal feedback in conversation is
provided next. The rest of the chapter exemplifies one possible approach to analysing
variation in the use of complex interactive discourse variables using corpus data. Section 3
provides an updated account of the classification framework for analyzing supportive verbal
feedback referred to above, followed by discussion of the methods and results of the gender
variation analysis in section 4.
2 Gendered speech styles and the use of verbal feedback
Research on gender and language use in Western, English-speaking societies has consistently
reported women to be generally more facilitative, cooperative conversationalists than men (e.g.
Aries 1976; Coates 1996; 1998, 2003, 2004; Crawford 1995; Holmes 1995, 2006; Holmes &
Meyerhoff 2003; Johnson & Meinhof 1997; Romaine 1999; Talbot 1998; Tannen 1990, 1993,
1994; Wertheim, Bailee & Corston-Oliver 1998; Wodak 1997). In particular, it has been
claimed that women are more active listeners in the sense that they are more likely to use
conversational strategies, including supportive verbal feedback or response tokens, which
support the smooth progress of an interaction and facilitate others’ opportunities to speak.
Much of the early work in this area focused on quantifiable gender differences in the use of
closed sets of linguistic devices which were assumed to generically index putative ‘competitive’
or facilitative’ conversational styles. To begin with, these styles were often positioned as polar
opposites, though, as discussed further below, the argument was increasingly made that in reality
such differences were indicative of a distributional gendered style continuum which could be
drawn on by both men and women in different settings (e.g. Cameron 1995, 1997; Bing &
Bergvall 1996; Stubbe et al 2001; Coates 2004). This early quantitative work largely predated
the era of widespread access to large computerised corpora, and thus typically involved the
manual coding and counting of formally defined categories of features such as interruptions or
minimal feedback tokens in relatively small, purposively collected sets of naturally occurring or
elicited interactions. One marker assumed to index the more co-operative interactive style
associated with women was a greater use of supportive minimal response tokens such as mhmm,
mm, and yeah.
Many studies (though not all) reported that women tended to use more minimal responses
than men to signal their involvement or to encourage others to contribute and keep talking
(e.g., Dittman 1972; Fishman 1983; Gilbert 1990; Hirschman 1974; Hyndman 1985; Leet-
Pellegrini 1980; Munro 1987; Nordenstam 1992; Roger 1989; Reid 1995; Strodtbeck & Mann
1956; Stubbe 1978; Zimmerman & West 1975). However, such analyses took little account of
potentially confounding variables such as text and activity type or topic (see below), or of the
often complex interrelationship between the range of possible forms and the different actions
they can be used to perform ‘online’ in the local context of the discourse. These studies were
also often flawed from a statistical standpoint due to the interactive nature of the variables
being examined. For example, the frequency of feedback responses produced is obviously co-
dependent on other variables such as the relative amount of talk produced by each speaker,
and the degree to which each speaker holds the floor. It is not surprising therefore that
subsequent research showed the relative frequency of minimal responses to provide at best
only a very crude measure of interactional supportiveness by listeners, whatever their gender.
Lane, Hilder, E.Vine, B.Vine, Holmes, Marra & Weatherall 2003; Stubbe & Holmes 1995, 2000; Holmes &
Stubbe 1992, 2003; H 2003; Holmes, Stubbe & Marra 2003).
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
3
From the early 1980s a number of researchers using a more action-oriented or functional analytic
framework began to identify patterns of gender difference in the use of open-ended categories of
facilitative verbal feedback. These included sentence completions, echoes, cooperative
overlapping and other longer comments or questions which appeared to indicate active listening
or intense involvement. Several studies found that such responses tended to occur in less formal
contexts where aone-at-a-time model of turn taking (c.f. Sacks, Schegloff & Jefferson 1974) is
not strictly adhered to by participants. For example, Edelsky (1981) demonstrated that American
women's participation in faculty meetings increased during collaborative floorsections of the
interaction, while Nordenstam (1992:84) reported that Swedish women talking to one another
used a more varied and elaborated range of backchannel signals which were often more
emotionally charged(e.g. det var ju fantastiskt that's incredible). Other researchers have
observed that women often prefer collaboratively constructed all-together-nowtalk, especially
when exchanging personal stories or gossiping with close friends (Coates 1989, 1998, 2004;
Pilkington 1994, 1998).
Researchers also began to put forward evidence suggesting women and men may use and
interpret simple minimal responses such as mhmm somewhat differently. For instance, Maltz &
Borker (1982) claimed that men were more likely than women to interpret a simple minimal
response as implying agreement, and Fishman observed that women in heterosexual couples
were “particularly skilled at inserting mms, yeahs, ohs and other such comments throughout
streams of talk rather than at the end, very precisely timed to fall between the breaths of the
speaker (1983:96). Reid Thomas (1995) noted a relationship between gender and the timing and
intonation of simple minimal responses, and a tendency for women's responses to be interpreted
more often than men's as meaning I'm listening versus I agree. Fellegy (1995) reported that
different usage or distributional rules were assigned to minimal responses by the Anglo-
American men and women in her sample: even though overall frequencies were very similar, the
women spread minimal responses more evenly throughout the speaker's utterance than the men,
who tended to load up their responses at the ends of syntactic units.
More recent work on language and gender has increasingly questioned the value of earlier
quantitative research which was conducted predominantly within a ‘gender difference’ paradigm.
This tended to be based on an essentialist assumption of consistent gendered differences in
language behaviour, regardless of the context and purpose of the communicative event, or the
locally occasioned contingencies of talk-in-interaction. Since the mid-1990s many language and
gender researchers (myself included) increasingly moved away from formally based quantitative
approaches to more nuanced qualitative analyses of how people use a range of interactional
resources to ‘perform’ or construct gender identities as they talk within gendered communities of
practice (e.g. Bergvall 1999; Bing & Bergvall 1996; Cameron 1997, 2003; Eckert & McConnell-
Ginet 1999, 2003, 2007; Freed 2003; Holmes & Meyerhoff 2003; Holmes & Stubbe 2003;
McElhinney & Mills 2007; Stubbe 1998; Talbot 2003; Weatherall 2002).
As Cheshire (2002:439) points out, not all women prefer affective meanings, or speak in a
cooperative speech style; and … those who do, do not always do so. The same applies to men
and their apparent preference for referential meanings; and to (other) patterns of sex
differentiation”. In other words, pre-determined social categories like ‘men’ and ‘women’ are
not simple homogeneous groupings, any more than other common intergroup comparisons.
The reality is that the communicative behaviours observed in any particular interaction result
from a complex interaction between the socio-cultural norms associated with the enactment of
various social roles and identities (of which gender may not be the most salient) and other
situated aspects of the encounter. The specific requirements of the context or type of talk in
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
4
which speakers are engaged are likely to influence the use of particular features at least as
much as an individual’s ‘performance’ of gender or other aspects of social identity, and intra-
sex differences in communicative behaviour are thus likely to be at least as great as any inter-
sex differences. Insofar as generalisations in terms of sex or gender are possible at all, therefore,
it is important to take account of overlaps between women’s and men’s behaviour as well as
any actual or presumed differentiation (Freed & Greenwood 1996).
The use of verbal feedback is highly contingent on the local context and the sequential
organisation of the interaction at the point where it occurs. While it is impractical to micro-
analyse each interaction in a corpus sample to determine the effect of context line by line,
more macro-level contextual indicators such as text type have frequently been commented on
in relation to gender and the use of feedback, and provide a useful starting point for exploring
the part played by the discursive context. For example, Coates (1989, 1996) found that the
frequency and discourse functions of feedback responses in women’s conversation varied
systematically according to whether the participants were engaged in interaction-focused
talk (where no one speaker has primary responsibility for the development of the topic), or
“information-focused” talk such as narrative or descriptive sequences with a single “topical
speaker (West & Garcia, 1988). Coates found that in information-focused talk, listener
responses were used less frequently but in specific ways, for example to signal agreement to a
topic shift, or to mark active agreement with a speaker's summing up. In these text types the
immediate involvement of the listener in the development of the topic is less. Moreover,
narratives in particular are characterised by long turns which need to be negotiated initially by
the speaker, but then require less regular localised feedback from the listener than is the case
in other more dialogic text types.
The topic being discussed has also been shown to influence the degree and nature of listener
involvement. For example in a New Zealand study comparing gossip and non-gossip texts,
Pilkington (1994, 1998) found that, when gossiping, both sexes shifted to a relatively ‘high-
involvement’ style, characterised in part by higher use of overtly supportive minimal responses
and cooperative overlaps. However, Pilkington found that men and women were not equally
likely to select the same feedback strategies, and also had different frequency baselines. For
instance, when gossiping, men showed a preference for supportive minimal responses to signal
high involvement, while women used more overlaps; however, the women's baseline rate of
cooperative overlap use in general conversation was equivalent to that of men engaged in gossip.
Earlier research has also consistently reported that men and women tend to discuss different
types of topics (Kalcik 1975; Aries 1976; Haas & Sherman 1982; Aries & Johnson 1983; James
& Drakich 1993; Tannen 1990; Coates 1989, 1996, 2003). In same-sex settings women tend to
talk more often about personal feelings and interpersonal relationships, and about intimate topics
and daily activities, whereas men tend to speak less about themselves, their feelings or their
relationships with others, and more about shared activities such as sports and hobbies. Because
such topics pose potential threats to both speaker's and listener's face, it is predictable that they
would be associated with a more frequent use of affectively supportive listener feedback and
other addressee-oriented pragmatic devices such as you know which function both to encourage
and support contributions by other participants in an interaction, and to convey solidarity and
shared understanding. Clearly, if women talk more than men about personal topics, and if such
topics as a rule generate relatively more high-involvement feedback regardless of listener gender,
then it is reasonable to predict that women in a representative corpus will produce a higher
proportion of explicitly supportive feedback than men (assuming otherwise comparable
discourse contexts).
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
5
More recent research in this area has tended to embrace a post-modernist or social constructionist
paradigm which problematises the use of predetermined ‘essentialist’ social categories such as
gender and the decontextualised classification and analysis of linguistic features. Nevertheless,
as Cameron (1995) points out, much of the earlier gender difference research was well-
conceived and carefully executed, with a pattern of results that has proved to be remarkably
robust over time. What has principally changed is the way in which such patterns are now
interpreted. As such, this earlier body of work continues to provide a useful and relevant
starting point for analysing gender patterns in corpus data, with the important proviso that the
complexities of precisely how discourse features such as listener responses function in context
must be taken fully into account when interpreting the results. This point is explored further in
the next section.
3 Developing a workable analytic framework
This section describes a taxonomy of supportive verbal feedback devices derived from an
exploratory analysis of 1800 listener responses in 22 dyadic conversations and radio interviews
from the WSC (see Stubbe 1996, 1998), and explains its underlying theoretical rationale. This
analysis built on the conclusions drawn from the research reviewed above; namely, that the
definition of what constitutes a supportive feedback device should include more extended
feedback like cooperative overlapping and sentence completions as well as simple minimal
responses, and that it would be instructive to look more closely both at the functions and patterns
of placement or distribution of feedback responses in relation to patterns of variation. The
resulting model was used to classify and quantify gender patterns in the use of listener
feedback for the subsample analysis described in section 4 below.
The data for the initial exploratory study came from the informal speech section of the WSC.
This constitutes 75% of the overall corpus, while 50% of the corpus is private face-to-face
conversations (DPC) collected using the ‘friend of a friend’ method. The WSC overall is
roughly representative in terms of gender: women contributed 52% and men 48% of the final
transcribed words, reflecting the New Zealand population balance. However, the WSC does
not include representative samples from particular ethnic groups other than the indigenous
Māori people. In respect to other social variables, there is a bias towards the 20-30 year age
band in the DPC data category, and the conversation participants are much more likely to be
university-educated than the general New Zealand population at the time (Holmes, Vine &
Johnson 1998).3 The sample analysed for the exploratory study comprised approximately 50,000
running words in total, with 44 participants representing two age bands (16-34, 40-55), two
ethnic groups (36 Pakeha, 8 Māori), and an even split of men and women. Because only audio-
taped data was available, non-verbal feedback was not studied.
3.1 Defining verbal feedback
Analysis in context of the 1800 listener responses identified in the exploratory analysis
highlighted the sheer complexity and range of the interactional resources available to listeners.
As with other complex discourse variables, listener responses are not bound to a closed set of
forms, and the ‘meanings’ or actions performed by these utterances are, by definition, co-
constructed and contingent on the local sequential context. This makes it challenging to
reliably categorise these phenomena for the purposes of a variationist discourse analysis of
3 Of individuals with completed background information sheets in WSC conversations, 32.9% had a university
degree, compared with 8.5% of the New Zealand population in the 1996 census.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
6
corpus data (see also Pichler 2010).
The following excerpt from the WSC exemplifies just a few of the interactional devices listeners
may draw on to actively involve themselves in a conversation while an interlocutor holds the
floor, and provides some indication of the complexity with which these devices function
interactively. It is also a canonical example of the collaborative high involvement listening
style often associated with women in the language and gender research of the late twentieth
century, as discussed above. (Supportive verbal feedback devices are shown in bold).4
(1) Supportive verbal feedback
DPC283 (extract from section of text not included in WSC).
Context: A woman tells her friend about a botched hair dying job
1 Mona: [laughs] so this is really silly 'cause the hair's white okay
2 //they muck it up so i end up red + in a couple of months i've got to go back\ +
3 Carol: /[laughs]\\
4 Carol: [laughs]: have it str-=/
5 Mona: /=they'd strip streaks of the red strip it right back to white but
6 don’t do it she said don't do the whole head
7 Carol: [laughs] streak it white //[laughs]\
8 Mona: /streak bits of it white\\ and then eventually the red will grow
9 out and fade and the white will dominate but it's going to take months
10 Carol: yeah
11 Mona: that's why i want to get it back to something close to normal and i've got my niece's
12 wedding to go to in //december\ so i'm going to go down in posterity=
13 Carol: /oh no\\
14 Mona: =in the family photographs as a //redhead [laughs]\
15 Carol: /[laughs] \\ oh dear + hmm
16 Mona: still mark thinks it's lovely
17 Carol: does he=/
18 Mona: /=oh yes
19 Carol: oohh // + oohh [quietly]: how could he:\
20 Mona: /i think it's a real sexist thing i think he thinks it makes me look\\
21 you know a lot younger and [laughs] therefore it might make him look a little bit younger
22 if i'm //with him i think that's the reasoning\
23 Carol: /[drawls]: oh gosh: \\
24 aren't they pathetic
25 Mona: [laughs] they are yes [laughs]
In this example, Carol listens sympathetically to Mona's account of a recent disastrous visit to the
hairdresser to have her hair dyed. As narrator, Mona mostly holds the floor, but Carol too is
actively participating in the conversation. At appropriate points in the narrative, she laughs along
with Mona (L3, L4, L7, L15); provides supportive minimal responses yeah (L10), oh no
(L13), oh gosh (L23), oh dear and hmm (L15); quietly interpolates longer ‘backchannel
comments- have it str- (L4), oohh + oohh how could he (L19); partially echoes Mona's words in
L5 “strip streaks …right back to white” with streak it white (L7); and asks facilitative
(rhetorical) questions which function as aligning evaluations - does he (L17), aren't they pathetic
(L24).
Without ever taking the floor as such, Carol thus indicates in various ways that the story has her
active attention and interest, and encourages Mona to continue telling it. At the same time, the
finely nuanced choices she makes about the type, placement and intensity of verbal feedback she
4 See Appendix for transcription conventions. All names used in examples are pseudonyms.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
7
provides add a complex layering of additional meanings. She can be heard to express her
sympathy with and understanding of Mona's plight, and humorously reinforces shared aspects of
their friendship and social identity (e.g. gender, age, life experience) by reflecting and extending
further Mona’s attitude of mock exasperation with her male partner. Mona in turn weaves
reciprocal feedback to Carol into her own narrative by projecting possible completions, echoing
or responding to Carol’s laughter and other supportive comments, often in overlap with Carol:
they'd strip streaks (L5), streak bits of it white (L8), oh yes (L18), i think it's a real sexist
thing...(L20), they are yes (L25). On hearing the excerpt, one is also struck by the rapidity and
precision timing with which these various listener responses are interpolated into the stream of
talk, including latched utterances5 (L4-5, L17-18), and the echoes and overlaps deployed by both
participants as they co-construct the discourse (e.g. L5-8, L19-20, L 22-23).
This example highlights some important observations about verbal feedback in conversation.
First, all supportive listener responses typically share a basic interactive function of aligning
with the current activity (Stivers 2008) in that they appear to ratify or otherwise contribute to
the talk of a current speaker(Zimmerman &West, 1975: 104) and confirm that speaker and
listener share a common frame of reference without threatening the current speaker's floor
holding. Neither minimal responses nor more extended backchannels are generally heard as
interruptions or as separate turns at talk (Coates, 1989, 2004; Edelsky 1981; Tannen 1984).
Second, listener responses may convey several levels of meaning simultaneously: minimal
responses in particular punch above their weightin terms of their interactional importance
(McCarthy 2003). For instance, as we saw in Example 1, supportive responses can provide an
explicit warrant for a long turn to continue (Coates 1989; Schegloff 1993) and may also function
to pass back the floor in cases where the speaker has yielded the turn, or as a polite means of
bookinga turn at talk (Stubbe 1994). “Topicalising utterances (West and Garcia 1988) or
news receipts(Maynard & Frankel 2006) such as really or oh yeah play an important part in
ongoing topic development and maintenance, while jointly constructed minimal response
sequences are regularly used to formally close down a topic (Filipi & Wales 2003; West 2006).
At a referential level, verbal feedback allows listeners to display indirectly their own knowledge
or competence within the domain of action to which the topic relates. It can also indicate the
listener's stance towards what is being said by indexing a wide range of epistemic meanings
from relative indifference or doubt, through simple affirmation to enthusiastic interest and
agreement (Bublitz 1988; Gardner 2001, 2005; Stenstrom 1994; Stubbe 1994). In addition,
listener feedback can be used to express supportive/affiliative relational meanings, ranging from
simple acknowledgment of an interlocutor’s feelings to explicit expressions of sympathy,
interest, surprise, or enthusiasm.
Third, the forms and functions of listener responses are not in a one-to-one relationship.
Listeners have a wide range of possible feedback devices to select from, and can produce,
combine and distribute these strategically in many different ways according to the context. The
precise ‘meaning’ or interactional effect of such utterances is thus subject to conversational
inference, and is highly contingent on how an interaction unfolds turn by turn.
Moreover, although feedback is a ubiquitous feature of conversation in all languages, the
specific meanings that may be inferred from different forms, baseline frequencies or
distribution of listener feedback and the levels or types of feedback considered appropriate in
particular contexts, have been shown to vary widely across cultures (see Stubbe 1998 for a
5 The term ‘latched utterance’ is used here in a structural sense to refer to turns which follow-on with no
discernible gap or intervening pause between utterances.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
8
more detailed discussion). For example, the typical Finn is a silent listener by comparison with
members of many other European ethnolinguistic groups: verbal feedback occurs
infrequently, and interruption and overlaps are generally considered unacceptable (Lehtonen
& Sajavaara 1985). Japanese speakers on the other hand are reported to have a relatively high
baseline for verbal backchannelling; this has been linked to the cultural value of omoiyari
which places great value on the creation and maintenance of harmonious interaction (Hayashi
1988; White 1989). Xudong (2008) studied the use of listener responses in Mandarin Chinese
and Australian English conversations, and found a number of differences in frequency, types
and the positioning of listener responses. Chinese speakers used fewer listener responses, used
more paralinguistic vocalic forms rather than lexical items, and were more likely to place their
listener responses during a turn. New Zealand Māori,6 in common with other Pacific cultures,
also place great importance on the creation of involvement and solidarity in informal discourse
(White & Watson-Gegeo 1990; Britain 1992); however, in contrast with Japanese interactions,
Māori conversationalists place a strong reliance on non-verbal cues and implicit reference in
preference to the use of explicit verbal feedback (Metge & Kinloch 1984; Britain 1992; Stubbe
1998; Holmes & Stubbe 2000).
Finally, it is important to note that verbal feedback does not always function supportively as
in this excerpt. Listeners have been reported to use both delayed feedback and minimal
responses rather than full turns at talk as strategies for avoiding further participation in a
conversation or to shut down discussion of a topic (Zimmerman & West, 1975; Fishman,
1983). Listener responses may also be used to indirectly signal impatience (Stivers 2004) and
disagreement or doubt (Stubbe 1994). Bublitz (1988:184) observes that because minimal
responses such as mm can be inserted almost anywhere in the stream of talk, they are also an
excellent device for “pretending to listen”, while extended sequences of simultaneous speech
may just as readily be intended (or perceived) to interrupt an interlocutor’s turn as to function
as a cooperative overlap or projected completion (Tannen 1990). Any analysis of verbal
feedback cannot therefore rely on formal features alone but must take account of the localized
contingent nature of spoken interaction.
3.2 The supportive feedback continuum
Three main dimensions of variation in how listeners provided supportive verbal feedback were
identified in the exploratory analysis:
1) the type of feedback device (defined by lexical, paralinguistic and interactional features)
2) the frequency and distribution of feedback responses
3) the placement of feedback in relation to the current speaker’s utterances.
These variables were observed to interact in complex ways as resources for producing a range
of meanings particular to the local discourse context. The overall conclusion was that listener
responses are most appropriately characterised as a relatively open class of multi-functional
utterances operating along a continuum or backchannel gradient (cf. Stenstrom 1994: 81)
reflecting the nature and degree of joint involvement in the interaction,7 the degree of explicit
referential or affective support provided by the listener, and the degree of social distance or
6 The Māori people are New Zealand’s original Polynesian inhabitants comprising about 15% of the population.
Pakeha is a Māori word used widely in New Zealand English to refer to people of European (mainly British)
descent; this is the dominant cultural/ethnic grouping in New Zealand.
7 More recently, Norrick (2011) has made the related observation that more “insistent” response tokens that
function as assessments or challenges (e.g. wow, really) will often elicit further responses in their own right,
unlike less “obtrusive” responses such as uh-huh, and proposes that listener responses could also usefully be
ranked along a scale of obtrusiveness or insistency.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
9
solidarity (Stubbe 1991, 1994, 1998). (See Figure 1).
FIGURE 1: The supportive verbal feedback continuum
SUPPORTIVE VERBAL FEEDBACK
One-at-a-time
Lower
involvement
No referential
content
Neutral affect
and/or stance
+ Distance
Keep going,
I’m listening,
I understand
NMR
SMR
CO
All-together-now
Higher
involvement
Explicit referential
content
Positive affect
and/or stance
+ Solidarity
I agree,
I sympathize,
I’m (very) interested
This provides a flexible system (both culturally and situationally relative), that allows listeners
to take an active role in influencing the topical development and interactional structure of a
conversation, as well as to convey a wide range of social and affective meanings. Available
responses range from providing the minimum amount of interactional feedback compatible with
sustaining the conversation, right through to the joint construction of the discourse. At the
affective and propositional levels, a listener response may simply be a neutral attention signal, or
it may carry a greater or lesser degree of positive referential or affective meaning, such as
confirming or agreeing with another speaker's statement of fact or opinion, or explicitly
expressing empathy and interest in what is being said. Such strategies can function
simultaneously to encourage and support contributions by other participants in an interaction,
and to convey solidarity and shared understanding.
Listeners are thus able to vary the type, placement and distribution of the feedback responses in a
way that allows them to meet their conversational goals in particular contexts. Particular devices
or forms are also multi-functional, both in the sense that they can perform different actions
according to the context, and also in their ability to function on more than one level
simultaneously. Because the precise meaning constructed by a given feedback device is always
subject to contextual inference, there can be no strict correspondence between form and function.
It is nonetheless possible to make some general observations about how the various devices
available to listeners tend to function in relation to one another. The supportive feedback
continuum as represented in Figure 1 provides a useful way of conceptualising this complex
interaction between the range of forms and functions observed in this data.
3.3 Types of feedback
The choices a listener makes between minimal or extended response forms on the one hand, and
between affectively or propositionally neutral or explicitly supportive responses on the other,
EXTENDED RESPONSES
MINIMAL RESPONSES
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
10
both have implications for how listener responses function in context. The analysis distinguished
three main categories of verbal feedback: neutral minimal responses (NMRs), explicitly
supportive minimal responses (SMRs) and cooperative overlaps (COs). These three types of
feedback are defined and exemplified next.
3.3.1 Neutral minimal responses (NMRs)
To the left of the continuum are brief minimal responses which can be heard as interactionally
supportive, but which are affectively and referentially neutral in tone. Typical functions of such
neutral minimal responses might include signalling attention, understanding, a willingness to
keep listening, or the negotiation of a topic shift. NMRs are basically heard as an
acknowledgement or continuation signal, although they may sometimes imply a secondary
meaning of tacit agreement or reserving judgement. Example 2 illustrates the kinds of responses
classified as NMRs: these are drawn from a very small closed set of lexical items such as mm,
uh-huh, yeah, okay and variants (c.f. Hirschman 1974; Fishman 1983; Schegloff 1982; Jefferson
1984). They sound prosodically and lexically unmarked, are characterised by mid to low pitch,
fairly level intonation and relatively low volume, and are usually placed at the boundary of a
syntactic or tone unit (i.e. a transition relevance point (TRP) as defined originally by Sacks et al,
1974). It is quite possible for a listener to utter such neutral minimal responses more or less
automatically and still appear attentive (Bublitz 1988).
(2) Neutral minimal responses (NMRs)
DPC317 1030-1055
Context: Two male colleagues discuss a television programme.
Tama: well er the catholic bishops' conference has made statements on homosexuality
and so on //and\ um have taken into account //i i know\
Eru: /mm\\ /mm mm\\
Tama: what er Māori opinion is for Māori catholics
Eru: mm
3.3.2 Supportive minimal responses (SMRs)
Along the continuum to the right are verbal responses which are still minimal in form, but which
are heard as more emotionally and epistemically charged. SMRs are more differentiated in
function or actions performed, expressing meanings such as sympathy, interest, surprise, and
explicit or enthusiastic agreement (c.f. Nordenstam 1992; Pilkington 1994, 1998; Stenstrom
1994; Stubbe 1994, 1998; Reid 1995). Listeners typically achieve this by clustering several
neutral minimal responses together in quick succession, or by using prosodically marked
minimal response forms which can be heard as being explicitly supportive in some way. SMRs
can be realised by a wide range of short utterances including explicit agreement markers and
minimal responses like mm, uh-huh, which are marked paralinguistically by extended pitch span,
raised voice range and higher volume and/or by rapid repetition. While SMRs also function in an
aligning way as an acknowledgement or continuer, they simultaneously signal an explicitly
supportive affective or positively valenced evaluative meaning of some sort along a continuum
of ‘strength’. Excerpt 3 provides an illustration of the forms and placement typical of this
response category.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
11
(3) Supportive minimal responses (SMRs)
DPC283 (extract from section of text not included in WSC)
Context: Two female friends discuss a botched hair-dying job
1 Mona: so whether I have to go back and get one of these colour bath things //done\ every four weeks=
2 Carol: /oh gee\\
3 Mona: =i don't know but i- and i don't know how much that will cost
4 Carol: yeah
5 Mona: so that's the annoying thing //all i'm asking them for is a refund\
6 Carol: /precisely ++ yeah\\
7 Mona: fo- for the original amount i //spent there\ and if they're not careful i'll take them to=
8 Carol: /exactly\\
9 Mona: =the small claims tribunal and ask for maintenance money as well
In this excerpt, Carol uses SMRs four times in quick succession to construct an affiliative stance,
to demonstrate her active attention to and interest in Mona’s story, and to signal her continued
willingness to be in the listener role as Mona continues with her extended narrative. In L1 she
expresses emphatic surprise with oh gee, produced in overlap. When Mona finishes her turn in
L3, Carol produces yeah with a similar intonation contour, which constructs it as an explicitly
affiliative response while at the same time functioning as a continuer. Mona then renews her
narrative thread, marking the transition to the upshot of her narrative with so. Carol’s next SMRs
in L6 (precisely ++ yeah) and L8 (exactly) are designed as explicit agreements with Mona’s
stated position, with the positively valenced lexical choices boosted by the prososdic features,
and in the case of L6, by the addition of the post-posed yeah.
3.3.3 Cooperative overlaps
Listeners may also signal high involvement or solidarity with more extended feedback of various
types, referred to here collectively as cooperative overlaps. This term, as used here, includes brief
interjections, sentence completions, echoes and repetitions, through to more extended segments
of simultaneous or actually overlapping speech which may include paraphrases, comments,
elaborations and questions.8 These usually align and/or affiliate with the content of the other
speaker’s utterance in some way, and are sometimes followed by an indication of support from
the primary speaker. Like minimal responses, cooperative overlaps seem to function along a
continuum of lesser to greater involvement in the primary speaker's turn. Features such as
sentence completions and echoes function much like SMRs in that they are overtly supportive,
yet carry little or no independent or new referential meaning. As Example 4 illustrates,
cooperative overlaps differ from minimal responses in that they are not drawn from a closed set
of forms, and listeners are more involved in the joint construction of text. At the high
involvement end of the scale co-operative overlapping becomes duetting(Coates 1991; Falk
1980) or contrapuntal talk which may appear to resemble disruptive interruptions, although
participants clearly do not orient to them as such.
8 Previous labels for this type of verbal feedback include "extended backchannels" (West & Zimmerman 1983),
non-turn-competitive incomings (French & Local 1983), non-floor-holding turns (Edelsky 1981), co-
operative overlaps(Tannen 1990), supportive interruptive forms(Stubbe 1991, 1994), secondary speaker
contributions (Bublitz 1988) and all-together-now” talk or interruptions (Coates 1989; Dunne & Ng 1994;
Pilkington 1994, 1998).
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
12
(4) Cooperative overlaps (COs)
DPC080:0430-0445
Context: Two female friends discuss proposed purchase of a piece of land.
Kate: mm i mean if you put your house up here you'd be tucked away from your property +
//you'd be tucked away from- yes it would actually i think + \
Tessa: /it would be good for vegetable gardens and things like that round there wouldn't it +\\
Kate: yeah=/
Tessa: /='cause it looked like very fertile soil there
3.5 Frequency, distribution and placement of feedback responses
In previous research, as noted earlier, the overall frequency of verbal feedback throughout an
interaction has often been assumed to relate directly to how supportive a particular listener is
being. However, the analysis of the WSC data suggested that at the macro-level at least, the
overall frequency of feedback in an interaction can at best provide only a crude measure of
interactive supportiveness (see also Stubbe 1994). This is because, in addition to the variability in
the type of feedback response already discussed, both the frequency and distribution of
feedback also varies both within and between particular interactions. This variation appears to
correlate with a number of different factors, including the text types produced by the participants,
the overall balance between participants in the roles of speaker and listener, and the nature of the
topics under discussion, as well constraints imposed by the immediate discourse context.
On the other hand, at a more localised discourse level, the way in which feedback responses are
distributed had much clearer functional implications in the WSC data. For example, a clustering
of responses in close succession appeared to be consistent with a higher level of involvement,
either at an affective level, or with some aspect of discourse structure, independently of the form
of the verbal response. Thus a cluster of NMRs at a point of topic shift could indicate the
listener’s alignment with the current speaker taking an extended turn. By way of contrast, the
more spaced out but constant and explicitly supportive feedback provided by Carol in examples
(1) and (3) also suggested high involvement and explicit positive affect, but in this case for the
reason that such regular feedback is not expected throughout a narrative sequence, and its
presence therefore carries relatively more functional weight. Examples such as these show that it
is not necessary or appropriate to use verbal feedback to the same extent or in the same way in all
contexts, and that speakers are sensitive to such norms. Therefore, although a high relative rate of
verbal feedback can be one indicator of conversational involvement and mutual interest, this
cannot be measured in absolute terms.
In addition, minimal responses often follow a consistent pattern of precise and well-timed
placement at the end of information units such as the tone group or clause (Coates 1989,
Dittman and Llewellyn 1967) or at phrasal junctures in predicates (Fellegy 1992). The latter
study also reported a pattern of sensitivity to semantic completion at an underlying level.
Intonation plays a part too in signalling where a minimal response is expected to occur in
certain text types such as narratives (Corston 1993). However, in the WSC data, as elsewhere,
minimal responses also frequently occurred at non-boundary points, a phenomenon which has
been associated with more active listening styles, as noted in 3.1 above. The link between the
contrapuntal placement of feedback responses and the listener's degree of involvement in the
conversation was seen most clearly in the case of cooperative overlaps.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
13
4 Analysis of gender variation in WSC sub-sample
4.1 Sample and method of analysis
A sample of eight dyadic conversations (roughly 17,288 running words in total) was selected
from the WSC. Each extract was approximately ten minutes in length, beginning and ending at
natural topic boundaries. Age and social class were controlled in selecting the sample: all 16
participants were between 40 and 60 years old, and came from professional backgrounds.
Further, the speakers in each dyad were matched for both gender and ethnicity, with two
conversations representing each gender/ethnicity group (i.e. Māori men, Māori women, Pakeha
men and Pakeha women, as shown in Table 1.
TABLE 1: Sample design
Pakeha/NZ European
Māori
Male
4 (2 dyads)
4 (2 dyads)
Female
4 (2 dyads)
4 (2 dyads)
In total, 905 instances of supportive interactive feedback were identified and classified as NMRs,
SMRs or COs. Given the interactive nature of listener feedback, a simple comparison of
individual raw scores would not allow meaningful comparisons between individuals or groups of
speakers. However, because participants were matched for gender and ethnicity, it was possible
to use the dyad as the statistical unit of analysis. (Nevertheless, it is important to note that this
does introduce a different limitation, namely that by using only matched dyads, the patterns of
interaction identified may not generalise to mixed dyads or groups).
The raw totals and the total number of words uttered for each dyad were used to calculate the
rate of feedback responses in relation to the total word count expressed as the number of
occurrences per 1000 words. The proportion of each type of feedback in relation to the total
number of feedback responses was then calculated for each dyad, expressed as a percentage
figure. The statistical significance of differences between sets of dyad index scores representing
each gender or ethnicity group was evaluated using Mann-Whitney U (for non-parametric, 2
sample score comparison).
The placement of minimal responses was also investigated. Minimal responses occurring at
transition relevance points (i.e. potential or actual points of speaker change) were included in the
count, along with the more frequent within-turn or backchannel responses which are
interpolated into the primary speaker's utterance. Instances of clustering of responses were noted,
and in the case of within-turnminimal responses, a count was done to determine the relative
frequencies of responses occurring at the boundaries of syntactic units or tone groups versus
those that were inserted at non-boundary points.
A macro-level analysis of the discourse structure of the eight extracts was then carried out, to
check the possible influence of contextual factors such as text type and topic of conversation
on the feedback patterns observed. The results of this analysis were subsequently cross-
checked against the wider dataset of 22 interactions from which the sample was drawn to
confirm the generalisability or otherwise of any patterns observed.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
14
4.2 Quantitative patterns
4.2.1 Overall frequency of supportive verbal feedback responses
Table 2 shows the total rate of occurrence per 1000 words of all types of verbal feedback
responses (calculated for both dyads in each gender x ethnicity category).
TABLE 2: Frequency of feedback responses (all categories)
(Rate per 1000 running words)
Male
Dyad
Rate
Dyad
Rate
Pakeha
PF1
60
PM1
55
PF2
65
PM2
82
Subtotal
63
69
Māori
MF1
38
MM1
59
MF2
34
MM2
42
Subtotal
36
51
TOTAL
Female dyads
50
Male dyads
60
In terms of gender, the overall averages for men and women were broadly comparable at 60
and 50 per 1000 words respectively overall, though the highest rate was produced by a Pakeha
male dyad (PM2), with the Māori women producing by far the lowest rate of feedback.
The data also shows a clear grading of feedback rates in this sample between Māori and
Pakeha, as shown in Figure 2.
FIGURE 2: Verbal feedback rates by gender and ethnicity
69
63
51
36
0
10
20
30
40
50
60
70
80
Pakeha
Male Pakeha
Fe male
Māori
Male
Māori
Fe male
Rate per 1000 words
This pattern is consistent with previous research on cross-cultural differences, and is at least
partly explained by the observation in previous research that Māori interactional norms
include a lower baseline for the amount of verbal feedback considered appropriate or
necessary, together with differences in the use and interpretation of pauses and silence and a
greater use of indirect feedback strategies such as code switching (see Metge & Kinloch 1984;
Britain 1992; Stubbe 1998). This result nicely illustrates the pitfalls in assuming that gender
or ethnicity somehow operate as unitary categories when we are analysing variation (see also
Stubbe 1998).
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
15
4.2.2 Distribution of different feedback types
Minimal versus extended feedback (MR vs CO)
The figures in Table 3 shows that, as expected, minimal responses (NMRs and SMRs combined)
accounted for the majority (74% to 88%) of all feedback responses produced by each dyad,
regardless of gender or ethnicity. However, the data also showed a small but consistent gender
difference across the sample, with females tending to produce proportionately fewer minimal
responses and more cooperative overlaps than males. The largest gap was between Māori women
and men, but the proportions for Pakeha and Māori women were comparable and in both cases
were higher than for both groups of men.
TABLE 3: MRs and COs as a proportion of all feedback
Male
% MR
% CO
% MR
% CO
Pakeha
74
26
80
20
ori
77
23
88
12
Overall
75.5
24.5
84
16
This pattern may well have affected the gender distribution for overall feedback rates reported
above because of the way feedback responses were indexed here (namely in terms of
occurrences per 1000 words). For the same reason, the figures shown in Table 3 for cooperative
overlaps are probably understated by comparison with those for minimal responses, and the
gender variation is therefore likely to be greater than is immediately apparent from these
statistics.
Neutral versus supportive minimal responses (NMR vs SMR)
Figure 3 demonstrates that, in addition to a higher proportion of minimal responses overall, the
male dyads in this sample consistently produced a higher proportion of neutral minimal
responses (NMRs) than the female dyads, and a correspondingly lower proportion of explicitly
supportive minimal responses (SMRs). The greatest proportion (65%) of female minimal
responses comprised SMRs, whilst the reverse was true of males, for whom NMRs comprised
60% of total minimal responses.
FIGURE 3: Proportion of minimal response types by gender
35
60
65
40
0
20
40
60
80
Fem ale Male
%
NMR
SMR
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
16
This inverse pattern of gender variation held true across the sample, although as we see in Table
4, the two Pakeha female dyads produced on average one third more SMRs than all other dyad
pairs in numeric terms.
TABLE 4: Frequency of SMRs by gender and ethnicity
Female
Male
Average rate per
1000 words
% total MRs
Average rate per
1000 words
% total MRs
Pakeha
30
64%
19
41%
Māori
19
68%
21
39%
Totals/Overall
%
24.5
65%
20
40%
Figure 4 plots the range of overall NMR rates for each dyad by gender and ethnicity. This clearly
shows that all male dyads produced higher NMR rates than all female dyads in this sample, a
result that was statistically significant (U (4,4) = 0, p=0.021).
FIGURE 4: Range of NMR rates per dyad (by gender and ethnicity)
31
23
22
40
9
15
11
12
0
5
10
15
20
25
30
35
40
45
Dyad 1
(Mā ori )
Dyad 2
(Mā ori )
Dyad 1
(Pakeha) Dyad 2
(Pakeha)
NMR Rate per 1000 words
Male
Female
Explicitly supportive feedback responses (SMRs vs CO)
When SMRs and cooperative overlaps (COs) were combined into a single functional category, a
much higher proportion of female feedback responses could be classified as explicitly
supportive, ‘high-involvement’ feedback than was the case for the men. Figure 5 shows that, on
average, almost three quarters (73.5%) of the women's feedback were explicitly supportive,
while just under half of the men's feedback came into this category. This result was statistically
significant (U (4,4) = 0; p= 0.021).
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
17
FIGURE 5: Proportion of explicitly supportive responses by gender
49.5
73.5
0
10
20
30
40
50
60
70
80
90
Male Female
%
Figure 6 clearly illustrates the interaction between gender and ethnicity in the use of explicitly
supportive feedback. Overall, the tendency to use more of this type of feedback is particularly
marked for the Pakeha women in the sample, as we also saw for the SMR rates shown in Table 4
above. However, because the proportions for the Māori listeners overall relative to Pakeha are
slightly lower anyway, the male-female differential remains roughly equivalent for both ethnic
groups.
FIGURE 6: Proportion of explicitly supportive responses by gender and ethnicity
46
53
65
82
0
10
20
30
40
50
60
70
80
90
100
Māori
Male Pakeha
Male
ori
Female Pakeha
Female
%
To sum up, overall the women in this sample tended to use a higher proportion of both types of
explicitly supportive verbal feedback. Interestingly, however, Māori and Pakeha women
signalled this higher level of involvement in different ways, with the Pakeha women making
proportionately greater use of SMRs, while the Māori women were more likely to use
cooperative overlaps.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
18
4.2.3 Placement of minimal responses within turns
The placement of minimal responses within a turn was the final variable to be quantified. All
minimal responses which occurred within a speaker's turn were classified according to whether
or not they occurred at the boundary of an information unit (a TRP/ transition relevance point).
Table 5 shows the proportions of within-turn minimal responses which occurred during a stream
of talk, rather than at a unit boundary.
TABLE 5: Proportion of non-boundary minimal responses within utterances
% of total MR rate
Female
Male
Pakeha
48
24
Māori
20
25
Overall %
34
25
The Pakeha women placed minimal responses at non-boundary points almost half the time, and
were also at least twice as likely to place minimal responses at non-boundary points compared
with either the Pakeha men or the Māori listeners of either gender. At one level this result is
consistent with the findings of previous research (e.g. Fishman 1983; Fellegy 1995; Coates 1991,
2004); namely, that women are more likely to place their responses throughout the stream of
talk, while men are more likely to wait for the end of a syntactic unit. However, it also once more
highlights the fact that gender patterns do not necessarily remain constant across other group
boundaries such as ethnicity.
The analysis revealed a particularly interesting pattern in relation to the use of NMRs and SMRs
in cases where a minimal response was placed throughout the stream of talk rather than at a
boundary point). Figure 7 shows that regardless of gender or ethnicity, a significantly higher
proportion of total MRs occurring in this environment are SMRs rather than NMRs.
FIGURE 7: Proportions of NMRs and SMRs at non-boundary points
61
36
28
25
25
21
18
15
0
10
20
30
40
50
60
70
Pakeha
female Pakeha
male
Māori
female
Māori
male
%
SMR
NMR
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
19
Again, this tendency was particularly marked in the case of the Pakeha women: while just under
half of their total minimal responses were placed at non-boundary points, over 60% of their
SMRs occurred there. This result correlates with the higher proportion of SMRs and COs used
by the Pakeha women overall, and is consistent with an overall tendency for both Pakeha and
Māori women to make greater use of contrapuntal responses as a feedback strategy.
4.2.4 Summary of quantitative results
The analysis revealed no significant correlation between gender and the overall frequency of
supportive listener responses. However, when the results were broken down further according to
the different feedback types and placement of responses within a turn, some interesting gender
patterns did emerge.
First, the men produced a significantly higher rate of neutral minimal feedback than the women,
and a higher proportion of their feedback overall was of the neutral and minimal response types.
By contrast, a markedly higher proportion of the women's feedback responses were overtly
supportive, with a correspondingly higher rate of cooperative overlaps; fewer were minimal
responses, and of those, a greater proportion were supportive minimal responses in non-
boundary positions. In short, the quantitative data showed a clear tendency for men to respond
more neutrally and minimally, while the women's feedback included a greater proportion of
responses which were both overtly supportive and more extended and contrapuntal in nature.
Second, there was a clear interaction in this data between gender and ethnicity: Māori and
Pakeha interlocutors were not equally likely to use the same strategies or to produce similar rates
of feedback. For example although both Māori and Pakeha women used a greater proportion of
overtly supportive feedback than men of either ethnicity, the Pakeha women produced such
feedback at a significantly higher rate than the Māori women. On the other hand, the gender
difference in the use of cooperative overlaps was more marked for the Māori listeners, while the
Pakeha women in this sample were the most likely of all the informants to place within-turn
minimal responses, particularly SMRs, at non-boundary points in an utterance.
4.3 Analysis of contextual variables
The degree of variation between dyads and within groups (men, women, Māori, Pakeha) in
the data analysed for this study suggests that the macro-contextual factors such as text type,
discourse structure (dialogic/monologic) and topic type, discussed earlier, may well have
influenced the feedback patterns reported in the previous section. Accordingly, a qualitative
analysis of the discourse structure of the eight extracts in the sub-sample was carried out to
identify such influences. The following tendencies were evident (and were also checked
against the 22-extract initial sample discussed in section 3 above).
Firstly, in those interactions where both speakers were actively involved in topical
development throughout all or most of the conversation, overall feedback rates tended to be
higher relative to those for other members of the same gender or ethnic group. These
interactions were also characterised by regular alternation of the role of 'topical speaker' with
both speakers contributing topic content more or less equally, and by a higher proportion of
opinion or problem-solving texts. Conversely, in more monologic interactions where one
speaker dominated in the role of topical speaker for relatively long stretches of the discourse,
and where the interaction comprised predominantly narrative and description text types,
overall feedback rates for that dyad were generally lower.
Secondly, as in previous studies, the characteristics of the topic(s) being discussed also correlated
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
20
with the degree and type of listener involvement. Regardless of gender, the highest rates of
explicitly supportive types of feedback occurred in those conversations, or sections of
conversations, where personal or sensitive topics were being discussed, especially where these
involved a degree of self-disclosure or, as in gossip or opinion texts, the overt evaluation of the
other person's behaviour or ideas.
These points will now be illustrated by close analysis of several extracts. This will serve at the
same time to demonstrate how the analysis and interpretation of distributional patterns observed
in corpus data can be enriched by taking contextual variables into account.
Example 5 is taken from an interaction between two Pakeha male speakers (Dyad PM2). This is
a typical excerpt from the conversation, throughout which both speakers contributed roughly
equal amounts and alternated in the role of topical speaker. The two men were colleagues in the
music industry; they knew each other well and were engaged in relaxed chat about topics of
mutual interest, and about which they shared a large amount of background knowledge. Their
conversation included a number of work-related gossip sequences, as well as a more general
sharing of ‘news and views’, and the excerpt illustrates how the two men both signalled their
close involvement as listeners through the use of frequent precision-timed verbal feedback.
(Verbal feedback responses are in bold).
(5) Two Pakeha men discuss the staging of a modern opera
DPC310:0225-0320
H: mm be interesting to see what [name of opera] is like
W: exactly
H: it's getting a huge amount of publicity i//suppose\just by virtue of it having been=
W: /mm\\
H: =picked up by the festival
W: yeah yeah
H: but er but it's sort of i mean i can't help//laughs\, noticing the comparison because
W: /laughs\\
H: i mean//it seems\ as if i mean it's on a similar scale and//and\ of course isn't it i=
W: /mm\\ /mm\\
H: =mean //those things\
W: /i don't know about\\ length it's probably quite a lot lon//ger (2) I suspect \
H: /is it do you think mm + mm\\
W: with chorus and
H: yeah
W: but i- i've never- i heard it in the last festival in that concert performance did you
H: no i was away +//mm\
W: /oh right\\ and the vocal writing was um i mean it just didn't sound as
though he knew where to put in +
H: yeah?
W: a higher i mean a high soprano singing quite fast narrative//about top g just absurd\
H: /[laughs] yeah\\
This dyad produced the highest overall rate of feedback of all the interactions in the sample, and
the third highest rate of overtly supportive feedback. It is also the only male interaction in the
sample where the topics discussed had any significant personal content, whereas three of the four
female interactions did. As predicted by previous research, the dominant text types (gossip,
discussion), discourse structure (dialogic) and topic types (personal, ‘office politics’) all correlate
here with a high rate of verbal feedback, in particular explicitly supportive feedback. However,
this predominantly takes the form of minimal feedback and laughter, not cooperative
overlapping.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
21
By comparison, the other Pakeha male interaction (Dyad PM1) produced a significantly lower
rate of feedback and was also noticeably different in other ways. Although the speakers here
also knew each other well, and the purpose of the talk too was primarily social, the nature of
the topic appeared to produce a number of differences at the discourse level. Here, the
participants were predominantly exchanging stories about their respective experiences with
painting their houses, a far more neutral topic, which generated mainly narrative and
description texts with one or other speaker holding the floor for extensive periods. In this
case, the choice of topic with its associated text types and turn taking structure appears to
have had a direct correlation with both the relatively low frequency of feedback, and the
predominant use of neutral minimal feedback.
A contextual analysis of the two Māori female interactions (MF1 and MF2) may provide
some explanation as to why these dyads produced the lowest rates of verbal feedback, both
overall and relative to the other Māori interactions (see Table 2 above), while at the same time
producing a higher proportion of SMRs and cooperative overlaps than any of the
conversations involving either the Māori or Pakeha men. Both of these interactions involved a
number of narrative sequences, and as we have seen, this text type tends to generate a lower
relative rate of verbal feedback. There is also an ethnic dimension - the use of various indirect
feedback devices was most marked in the interactions between the Māori women. These
included supportive silences, code-switching into Māori, deferred feedback, and implicit
contextualisation devices such as the second story phenomenon described by Sacks
(Jefferson 1992) where a participant retells and or adds to an interlocutor’s previous narrative
(discussed in greater detail in Stubbe 1998: 273-8). There were also, however, several
extended sequences within these two interactions where the participants engaged in very
active co-construction of the text, coinciding with a shift to a more emotive topic. The overall
count, based on each identified instance of localised verbal feedback across the extracts as a
whole, therefore downplays the actual amount of facilitative feedback going on in these two
interactions; this is reflected instead in the high proportion of explicitly supportive and
contrapuntal feedback in these interactions.
Finally, it is instructive to compare two interactions (MM1 and PF2) which are quite similar in
terms of their overall discourse structure and with middling overall rates of feedback (59 and 65
per 1000 respectively). In both cases the participants were friends who were attempting to reach
a consensus on a topic in which they had a strong interest, yet about which they had differing
opinions. Both conversations took the form of animated discussions, essentially dialogic in
structure, with each participant contributing very actively throughout the discussion. In both
cases, neither speaker can be said to have had any special claim to expertise, except where they
resorted to personal experiences to support a point. They could be observed to have strongly held
opinions and supporting evidence they wished to contribute, while at the same time carefully
managing the interaction to avoid making potential areas of disagreement overt. Based on
previous research, in such a context we might expect a relatively high rate of verbal feedback
overall, coupled with a greater incidence of explicitly supportive feedback, and indeed, both
interactions provide evidence of a relatively high degree of active listener involvement.
However, the way in which that involvement was enacted is quite different in each case.
In Example 6 (Dyad MM1), the two men are talking socially over morning tea at their
workplace, discussing a controversial television programme about the place of homosexual men
in traditional Māori society. One has seen the programme; the other has not, although he has
previously talked about it at length with his wife.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
22
(6) Two Māori men discuss a controversial television programme
DPC317:0325-0420
Tama: .and they- they still need to- they- they had a right to be part of the hapu //the\ kinship //group\
Eru: /yes\\ /yes\\
Tama: um and you see the normality as i see it from where i come from in- in my community was 'cause
you know we had people with that s- with those TRAITS +
Eru: yes yes
Tama: er they were not tre- if they were men
Eru: mm
Tama: they were given MEN'S work
Eru: yes yes
Tama: and you know [voc] and i know of a family that- that expected the- the- th- the person
with those showing those tendencies to be he was made to plough
Eru: mm
Tama: like the other brothers
Eru: yes
Tama: and he was expected to milk the cows like the //other\ brothers //+ \ expected to go fencing
Eru: /mm\\ /mm\\
Tama: like the others he was not given treatment other than what his gender //+\ required him to do
Eru: /yes\\
Eru: yes yes
Throughout this interaction, one participant often took the role of primary speaker for relatively
long periods, as in the example above, and although a relatively high proportion (33%) of within-
turn minimal responses occurred at non-boundary points, virtually none of the feedback provided
comes from the cooperative overlap category. The turn-taking organisation is thus very much of
the ‘one at a time’ type. Secondly, less than half of the feedback in this interaction can be
classified as being explicitly supportive in function. However, at a rate of 59 feedback responses
per 1000 words, this interaction has a notably higher rate of verbal feedback than any of the other
three Māori dyads. This interaction also does not exhibit the periodic monologues and long
pauses with no feedback which seem to account most obviously for the lower feedback rates
elsewhere in the Māori data (see Stubbe 1998:281). By comparison with more typical Māori
listener behaviour, these characteristics indicate relatively active and overtly supportive listener
involvement. It seems that explicit verbal feedback was a more important strategy here in
achieving the required balancing act than it was in the other Māori interactions which consisted
primarily of narrative and description text types, and probably functioned here largely as a
negative politeness device.
Excerpt 7 (Dyad PF2) provides a contrasting example. This was a discussion between two
women of a collection of site plans relating to a proposed joint land purchase, and provides a
marked contrast in terms of feedback use. It also provides an excellent illustration of how
participants can use both cooperative overlaps and non-boundary minimal responses to signal
how closely they are following the other speaker's thought processes. The excerpt below is
representative of the whole conversation, with the two women alternating in a series of short
turns, which often overlapped. The construction of the discourse was collaborative to such an
extent that it was often impossible to assign the role of topical or primary speaker to either
participant, and almost 90% of the feedback provided by both participants consisted of SMRs or
cooperative overlaps..
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
23
(7) Two Pakeha women discuss landscaping plans
DPC080:0800-0865
A: and if you can make rooms vistas to the lake and all sorts of things i mean it
could be just lovely couldn't it
B: or instead of having one big lake=
A: =yes //i would have s- a chain of little ones\
B: /you could have + a chain of \\ ponds and a walk
//among them or something like that like that a-\=
A: / mm yes up to about here up- up here \\
B: =chain of ponds
A: yep
B: like that ++ with //largish trees on the-\ on the intersections of the //walks\
A: /yes mm \\ /mm\\
In these two cases, a relatively high rate of minimal responses signalled high involvement for the
men (they were at the top of the range for the Māori participants), while the women used a
different set of strategies to achieve the same interactional goal (which, as pointed out earlier,
also pushed their overall frequency count down, because they used more extended and
contrapuntal sequences to provide their feedback).
The contextual factors described above help to account for the within-group differences in
overall rates of feedback reported above for individual dyads. It is also clear that although a
high relative rate of verbal feedback is one indicator of conversational involvement and
mutual interest, this cannot be measured in absolute terms. The appropriate amount, type and
placement of feedback varies in a consistent fashion according to both the context, topic and
structure of the interaction, as well as to the norms of the social group(s) to which the
participants belong, and any patterns of gender difference must be interpreted against this
complex backdrop.
5 Conclusion
5.1 Methodological issues
This chapter has highlighted several methodological issues which arise when we try to use
naturalistic corpus data to analyse patterns of occurrence for a complex discourse phenomenon
such as verbal feedback. These include the technical limitations and theoretical issues inherent in
attempting to ‘code and count’ when dealing with open-ended classes of pragmatic devices, the
problem of dealing in an analytically robust way with the absence of a one-to-one relationship
between form and pragmatic function, and the difficulty of teasing out the complex
interrelationships between the various factors that may correlate with or account for patterns of
inter-group variation.
The technical issues arising from trying to apply corpus linguistic approaches to the study of
discourse have been increasingly commented on by discourse and pragmatics researchers in
recent years as technical progress has allowed linguistic corpora and the associated analytic tools
to evolve into ever larger and more sophisticated forms. For example, Pichler argues that “a
uniform model of discourse variation” is needed which defines discourse variables more to more
adequately take account of their flexibility and multifunctionality (2010:1), and Wong & Peters
are rightly critical of the way in which “borrowing of terminology and a reliance on axiomatic
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
24
definition has resulted in a diverse nomenclature and an indeterminate inventory of forms”
(2007:1). These difficulties will only multiply further as corpus linguistics expands into the
realm of multimodal analysis of discourse (Knight & Adolphs 2007).
In the study presented above some of these issues were addressed by looking specifically at how
verbal feedback is used from a broad functional perspective, rather than by simply quantifying a
limited range of forms and plotting their distribution according to a pre-determined variable such
as sex of speaker, as has often been the case in the past. The approach taken here has highlighted
the sheer complexity and range of the interactional resources available to listeners. As we have
seen, verbal feedback devices are typically multi-functional and are characterised by a very fluid
relationship between the form used and the function or action performed. Listeners can be
observed to make sophisticated choices as to the type, placement and frequency of the responses
they provide in a way that is tailored to the emerging discourse context. This allows them to take
an active role in influencing the topical development and interactive structure of a conversation
as well as to convey a wide range of social and affective meanings.
The same characteristics which make verbal feedback such a flexible resource for listeners in
conversation also make the task of interpretation an extremely challenging one for the analyst.
The use and interpretation of verbal feedback is heavily contingent on the local sequential
environment, and what is ‘appropriate’ is thus constantly being (re)negotiated throughout a
conversation. Moreover, various other situational and contextual factors such as topic and text
type, the structure of the discourse and the relationship between the participants also correlate
and interrelate in complex ways with overall rates of feedback and the particular strategies
selected for use in any given interaction. Samples of corpus data cannot be expected to control
for this type of variability, but such factors will nonetheless inevitably have influenced what
might be expected from listeners at a given point in a conversation and therefore what they
actually did and how it might be interpreted.
Close attention to the local context is the key to understanding what listeners do and why they
might be doing it that way in a particular instance. Any apparent inter-group patterns of variation
must therefore always be interpreted in the light of the shared contextual and interpretive
frameworks demonstrably being oriented to by all participants in a conversation. In the case of
the present study, both men and women were observed to use a similar set of interactive
resources to indicate their degree of involvement in a conversation along a supportive feedback
continuum, and they oriented to similar normative expectations in relation to other variables such
as text type and topic. However, the ways in which they deployed these resources did appear to
vary systematically according to both gender and ethnicity.
The somewhat different patterns for the Māori and Pakeha women and men in this study also
underline the importance of not simply assuming that gender differences, or any other inter-
group differences for that matter, will apply across the board: ‘women’ and ‘men’ are not
homogeneous groups (c.f. Coates 1989:11) any more than people from a particular ethnic or
social group. Moreover, other personal and intergroup identities are bound to interact with
gender in complex ways, although it is interesting that, at least in the data analysed here, there
were nevertheless certain gender patterns which seemed to cross inter-group boundaries.
The listener's interactive and relational involvement in a conversation is the concept which linked
all these variables and underpinned the model of supportive verbal feedback used in the analysis
of gender variation presented in this chapter. According to this model, the interactive functions of
feedback can range from providing the minimum amount of interactive feedback compatible
with sustaining the conversation, to significantly influencing topic and turn taking structures,
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
25
right through to the joint construction of the discourse. At the propositional, relational or
affective level, listener feedback may involve a quite neutral attention signal, or it may carry a
greater or lesser degree of positive affective meaning, such as confirming or agreeing with
another speaker's statement of fact or opinion or explicitly expressing empathy and interest in
what is being said. Such strategies can function simultaneously to encourage and support
contributions by other participants in an interaction, and to convey solidarity and shared
understanding.
5.2 Feedback and gender
Despite the complexity of the data and the underlying pragmatics, this analysis nonetheless
revealed some clear correlations between gender and patterns of feedback use. The key finding
of the present study was that the males as a group provided proportionately more feedback which
was both affectively neutral and minimal in form, while the females provided a significantly
greater proportion of overtly supportive feedback, together with more extended and contrapuntal
responses of various kinds. However, the actual rates and localised distribution and placement of
feedback in this sample correlated more strongly with culturally relative norms and contextual
factors such as text type and topic than with gender.
These findings are consistent with previous research which suggests that men and women (or
different ethnic groups) (i) may use different strategies to achieve similar effects such as
signalling greater or lesser involvement in a conversation, or conversely, may use certain ways
of providing feedback to signal somewhat different meanings; and (ii) may have different
baselines or starting points with respect to the amounts and types of verbal feedback which are
considered appropriate in different contexts.
While at one level such observations are helpful in explaining the gender differences in listener
behaviour reported here, they do not address the deeper question of why it is that women and
men should have different response styles in the first place. Cameron (1995:41) argues that
differences in the ways women and men interact are socially produced, and have their roots in
unequal gender relations. Thus, if women and men systematically use different discourse
strategies, this is because they are engaged in fundamentally different tasks. She goes on to
suggest that different verbal practices should be seen simply as one aspect of the process by
which a particular community defines masculinity and femininity. Rather than seeing male and
female styles as simple markers of gender identity, she concludes that "...the styles themselves
are produced as masculine and feminine, and that individuals make varying accommodations to
those styles in the process of producing themselves as gendered subjects" (43).
Eckert and McConnell-Ginet take a similar approach in advocating that the study of language
and gender should be grounded in "...detailed examination of the social and linguistic activities
of specific communities of practice" (1995:469). Thus patterns of variation often emerge from
the co-construction of gender and other variables such as class or ethnicity (501), and when
speakers use a particular variant or speech style these do not simply reflect "a meaning already
set and waiting to be recycled", but simultaneously produce a social meaning (503).
Such an approach provides a useful framework for understanding the sometimes elusive patterns
of gender difference in a discourse feature as complex and variable as verbal feedback, one
which typically defies attempts to find clear-cut bipolar oppositions once the data is analysed in
context and in finer detail. Listeners can be observed to select from a wide range of feedback
strategies in order to influence the topical development and interactive structure of a
conversation, as well as to convey a range of social and affective meanings in a way that is
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
26
appropriate to the local discourse context. Moreover, localised verbal feedback of the kind
analysed here is only one aspect of the listener's repertoire: there are many other ways to signal
conversational involvement as well. It is therefore unlikely that there is any kind of
predetermined relationship between a social category such as sex/gender and patterns of
feedback use. A more plausible interpretation of the gendered patterns reported in this study,
therefore, is that verbal feedback functions as one of a number of interactive resources available
for doing genderas one aspect of identity work, in addition to its many other interactive and
social meanings.
In conclusion, this study has highlighted some of the limitations inherent in using linguistic
corpora to study gender variation in the use of a complex discourse feature such as supportive
verbal feedback. As we have seen, the multi-functionality and highly contingent nature of
listener responses makes the job of quantifying and interpreting these discourse variables in
natural conversations an extremely challenging one. One way of dealing with this challenge is by
combining quantitative and qualitative approaches to the data, and by taking due account of the
relevant contextual variables and functional complexities in interpreting the results. In the study
presented here, the verbal feedback continuum provided a unifying analytic framework to
account for the complex and fluid interrelationships between the various forms and functions of
supportive verbal feedback. Using a flexible analytic model such as this makes it possible to
generate valuable insights into wider distributional patterns which complement more in-depth
micro-analysis of smaller data sets. Notwithstanding the very real challenges, linguistic corpora
therefore remain an important and useful resource for the study of interactive discourse variables.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
27
Acknowledgements
1 The research on which this chapter is based was made possible by a grant from the New
Zealand Foundation for Research, Science and Technology, and was carried out during the
author’s tenure as research fellow with the School of Linguistics and Applied Language
Studies, Victoria University of Wellington from 1996 - 2003.
2 I am grateful to all those who allowed their interactions to be recorded and analyzed as
part of the Wellington Corpus of Spoken New Zealand English. I also thank Robert Sigley
for his assistance with the statistical tests, and David Britain, Chris Lane, Janet Holmes
and an anonymous referee for their useful input on earlier iterations of this work. Any
errors and infelicities remain entirely my own responsibility.
Transcription conventions
The conventions used in the examples in this paper are a simplified version of those developed
for the Wellington Corpus of Spoken New Zealand English.
(hello) Transcriber's best guess at an unclear utterance
? Rising or question intonation
- Incomplete or cut off utterance
YES Capitals denote emphatic stress
[laughs] [softly] Paralinguistic features or transcriber comment
+ Pause of up to one second
//... ....\ Simultaneous speech
/.... ....\\
…= Current speaker’s utterance continues onto another line
=…
…=/ Latching (no discernible gap between utterances)
/=…
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
28
References
Aries, E. J. 1976. Interaction patterns and themes of male, female and mixed groups. Small Group
Behaviour, 7, 1: 7-18.
Aries, E. J. and F. L. Johnson 1983. Close friendship in adulthood: conversational content between same-
sex friends. Sex Roles, 9, 12: 1183-1196.
Bergvall, V.L. 1999. Toward a comprehensive theory of language and gender. Language in Society,
28, 2: 273-293.
Bing, J.M. & V.L. Bergvall 1996. The question of questions: beyond binary thinking. In
V.L.Bergvall, J.M. Bing & A.F. Freed (eds.), Rethinking Language and Gender Research: Theory
and Practice. New York: Longman, 1-30.
Britain, D. 1992. Linguistic change in intonation: The use of high rising terminals in New Zealand
English. Language Variation and Change 4: 77-104.
Bublitz, W. 1988. Supportive Fellow-Speakers and Co-operative Conversations: Discourse Topics and
Recipient Actionin a Particular Type of Everyday Conversation. Amsterdam: Benjamins.
Cameron, D. 1995. Rethinking language and gender studies: some issues for the 1990s. In Mills, S. (ed.)
Language and Gender: Interdisciplinary Perspectives. London and New York: Longman. 31-44.
Cameron, D. 1997. Performing gender identity. In S. Johnson & U.H. Meinhof (eds.), Language and
Masculinity. Oxford: Blackwell, 47-64.
Cameron, D. 2003. Gender and language ideologies. In J. Holmes and Miriam Meyerhoff (eds.)
Handbook of Language and Gender. Blackwell: 447-467.
Carter, R. and S. Adolphs 2008. Linking the verbal and visual: new directions for corpus linguistics.
Language and Computers, 64:1, 275-91.
Cheshire, J. 2002. Sex and gender in variationist research. In J.K. Chambers, P. Trudgill and N.
Schilling-Estes (eds.) Handbook of Language Variation and Change. Oxford: Blackwell, 2002, pp.
423-43.
Coates, J. 1987. Epistemic modality and spoken discourse. Transactions of the Philological Society
1987: 110-131.
Coates, J. 1989. Gossip revisited: language in all-female groups. In J. Coates and D. Cameron (eds.),
Women in their Speech Communities. U.K.: Longman. 94-122.
Coates, J. 1991. Women's cooperative talk: a new kind of conversational duet? In C. Uhlig and R.
Zimmerman (eds.) Proceedings of the Anglistentag 1990 Marburg. Tubingen: Max Niermeyer Verlag.
Coates, J. 1996. Women Talk. Oxford: Blackwell.
Coates, J. 1998 (ed.). Language and Gender. A Reader. Oxford: Blackwell.
Coates, J. 2003. Men Talk. Oxford: Blackwell.
Coates, J. 2004. (3rd Ed). Women, Men and Language. A Sociolinguistic Account of Gender
Differences in Language. UK: Pearson Education.
Crawford, M. 1995. Talking Difference: On Gender and Language. London and Thousand Oaks:
Sage.
Dittmann, A.T. 1972. Developmental factors in conversational behaviour. Journal of Communication 22:
404-423.
Dunne, M. and S.H. Ng 1994. Simultaneous speech in small-group conversation: all-together-now and
one-at-a-time? Journal of Language and Social Psychology, 13, 1: 45-71.
Eckert, P. and S. McConnell-Ginet 1995. Constructing meaning, constructing selves. Snapshots of
language, gender and class from Belten High. In K. Hall and M. Bucholtz (eds.) Gender Articulated.
Language and the Socially Constructed Self. New York and London: Routledge. 469-507.
Eckert, P. and S. McConnell-Ginet 1999. New generalizations and explanations in language and
gender research. Language in Society, 28, 2: 185-201.
Eckert, P. and S. McConnell-Ginet 2003. Language and Gender. Cambridge: Cambridge University
Press.
Eckert, P. and S. McConnell-Ginet 2007. Putting communities of practice in their place. Gender and
Language 1,1: 27-37
Edelsky, C. 1981. Who's got the floor? Language in Society 10: 383-421. Reprinted in D Tannen (ed.)
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
29
1993 Gender and Conversational Interaction. Oxford: Oxford University Press.
Falk, J. 1980. The conversational duet. Proceedings of the 6th Annual Meeting of the Berkeley
Linguistics Society. 6: 507-514.
Farr, F. 2003. Engaged listenership in spoken academic discourse: the case of student-tutor meetings.
Journal of English for Academic Purposes. 2, 1: 67-85.
Fellegy A. M. 1995. Patterns and functions of minimal response. American Speech. 70, 2, 186.
Filipi, A. and R. Wales 2003. Differential uses of okay, right, and alright, and their function in
signaling perspective shift or maintenance in a map task. Semiotica 147, Nov 2003.
Fishman, P. 1983. Interaction: the work women do. In B. Thorne, C. Kramarae and N. Henley (eds)
Language Gender and Society. Rowley, Massachussets: Newbury. 89-102.
Freed, A. 2003. Epilogue: Reflections on language and gender research In J. Holmes and M.
Meyerhoff (eds.) Handbook of Language and Gender. Blackwell: 699-721.
Freed, A.F. and A. Greenwood 1996. Women, men and type of talk: What makes the
difference?’Language in Society 25, 1-26.
French, P. and J. Local 1983. Turn-competitive incoming. Journal of Pragmatics 7, 1: 17-38.
Gardner, R. 1997. The listener and minimal responses in conversational interaction. Prospect, 12, 12-
32.
Gardner, R. 1998. Between speaking and listening: The vocalisation of understandings. Applied
Linguistics, 19, 204–224.
Gardner, R. 2001. When Listeners Talk: Response Tokens and Listener Stance. Amsterdam: John
Benjamins Publishing Company.
Gardner, R. 2007. The Right connections: Acknowledging epistemic progression in talk.Language in
Society 36, 319–341
Gilbert, J. 1990. Secondary School Students talking about science: language functions, gender and
interaction in small group discussions. Unpublished MA Thesis. Wellington: Victoria University.
Haas, A. and M. Sherman 1982. Reported topics of conversation among same-sex adults.
Communication Quarterly, 30, 4: 332-342.
Hayashi, R. 1988. Simultaneous talk - from the perspective of floor management of English and
Japanese speakers. World Englishes 7: 269-288.
Hirschman, L. 1974. Analysis of assertive and supportive behaviour in conversations. Paper presented at
Linguistic Society of America Meeting, July 1974.
Holmes J. and M. Meyerhoff 2003 (eds.). The Handbook of Language and Gender. Oxford:
Blackwell.
Holmes, J. 1995. Women, Men and Politeness. London and New York: Longman.
Holmes, J. 2006. Gendered Talk at Work: Constructing Gender Identity Through Workplace Discourse.
New York and Oxford: Blackwell.
Holmes, J. and M. Meyerhoff 2003. Different voices, different views: an introduction to current
research in language and gender. In J. Holmes and M. Meyerhoff (eds.) Handbook of Language
and Gender. Blackwell: 1-17.
Holmes, J. and M. Stubbe 1992. Women and Men Talking: Gender-based Patterns of Interaction. In
S. Olsson (ed.) The Gender Factor. Dunmore Press 1992, 149-163.
Holmes, J. and M. Stubbe 1997. Good listeners: gender differences in NZ conversation. Women and
Language Vol XX, No. 2, 7-14.
Holmes, J. and M. Stubbe 2003. Discourse in gendered workplaces: how do women manage it? In J.
Holmes and M. Meyerhoff (Eds) Handbook of Language and Gender. Blackwell: 573-600.
Holmes, J., B. Vine and G. Johnson 1998. The Wellington Corpus of Spoken New Zealand English: a
Users’ Guide. Wellington: School of Linguistics and Applied Language Studies, Victoria
University of Wellington.
Holmes, J., L. Burns, M. Marra, M. Stubbe and B.Vine 2003. ‘Women managing discourse in the
workplace.’ Women in Management Review 18 (8) 2003, pp 414- 424
Holmes, J., M. Stubbe and M. Marra 2003. Language, Humour and Ethnic Identity Marking in New
Zealand English. In C. Mair (ed.) The Politics of English as a World Language. New Horizons in
Postcolonial Cultural Studies. ASNEL Papers 7. Amsterdam – New York: Rodopi, pp 431 – 456.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
30
Hyndman, C. 1985. Gender and language differences: a small study. Unpublished terms paper.
Wellington: Victoria University.
James, D. and J. Drakich 1993. Understanding gender differences in amount of talk: a critical review of
research. In D. Tannen (ed.) Gender and Conversational Interaction. New York, Oxford: Oxford
University Press. 281-312.
Jefferson, G. 1984. Notes on a systematic deployment of the acknowledgment tokens ‘yeah’ and
‘mmhm’. Papers in Linguistics 17 (2): 197–216.
Jefferson, G. 1993. Caveat speaker: Preliminary notes on recipient topic-shift implicature. Research
on Language and Social Interaction, 26, 1-30.
Jefferson, G. 2002. Is ‘no’’ an acknowledgment token? Comparing American and British uses of
(+)/(-) tokens. Journal of Pragmatics 34 , 1345–1383
Johnson, S. and U. H. Meinhof 1997 (eds). Language and Masculinity. Oxford: Blackwell.
Kalcik, S. 1975. ‘Like Ann's gynecologist or the time I was almost raped. Personal narratives in
women's rap groups. Journal of American Folklore, 88: 3-11.
Knight, D & S. Adolphs 2008. Multi-modal corpus pragmatics: The case of active listenership. In J.
Romero-Trillo (ed) Pragmatics and Corpus Linguistivs. A Mutualistic Entente, 175-90. Berlin. New
York: Mouton de Gruyter, 175-90.
Leet-Pellegrini, H. M. 1980. Conversational dominance as a function of gender and expertise. In: H.
Giles, P. Robinson and P. Smith (eds) Language: Social Psychological Perspectives. Oxford:
Pergamon Press, 97-104.
Lehtonen, J. and K. Sajavaara, 1985. The silent Finn. In: D. Tannen and M. Saville-Troike (eds.)
Perspectives on silence. Norwood, N J: Ablex, 193-201
Maltz, D.N. and R.A. Borker 1982. A cultural approach to male-female miscommunication. In J.
Gumperz (ed.) Language and Social Identity. Cambridge: Cambridge University Press, 196-216.
Maynard, D.W. and R.M. Frankel 2006. On diagnostic rationality: bad news, good news, and the
symptom residue. In J. Heritage and D.W. Maynard (eds.) Communication in Medical Care.
Interaction between Primary Care Physicians and Patients. Studies in Interactional Sociolinguistics
20. Cambridge: CUP, 248-278.
McCarthy, M. 2003. Talking Back: Small Interactional Response Tokens in Everyday
Conversation. Research on Language and Social Interaction, 36(1), 33–63
McElhinny, B. and S. Mills 2007. Editorial: Launching Studies of Gender and Language in the Early
21st Century. Gender and Language 1,1: 1-13.
Metge, J. and P. Kinloch, 1984. Talking past each other: Problems of cross-cultural communication.
Wellington: Victoria University Press/Price Milburn.
Munro, F. 1987. Female and male participation in small-group interaction in the E.S.O.L. classroom.
Unpublished terms project. Graduate Dilpoma in TESOL. Sydney: Sydney College of Advanced
Education.
Nordenstam, K. 1992. Tag questions and gender in Swedish conversations. Working Papers on
Language Gender and Sexism. 4, 1: 69-77.
Norrick, N. 2011 (In Press). Listening practices in English conversation: The responses responses
elicit. Journal of Pragmatics (2011), doi: 10.1016/j.pragma.2011.08.007
Orestrom, B. 1983. Turn-taking in English Conversations. Lund: Liber.
Pichler, H. 2010. Methods in discourse variation analysis: Reflections on the way forward. Journal of
Sociolinguistics. 14: 581-608.
Pilkington, J. 1994. Women, Men and Gossip: What's the Story? Unpublished MA Thesis. Victoria
University of Wellington.
Pilkington, J. 1998. ‘Don’t try and make out that I’m nice!The Different Strategies women and Men
Use when Gossiping. In Coates, J. (ed.). Language and Gender: A Reader. Oxford: Blackwell,
254-269.
Reid Thomas, H. 1995. The use and interpretation by men and women of minimal responses in informal
conversation. Working Papers on Language Gender and Sexism. 5, 1: 65-89.
Reid, J. 1995. A study of gender differences in minimal responses. Journal of Pragmatics. 24: 489-512.
Roger, D. 1989. Experimental studies of dyadic turn-taking behaviour. In D. Roger and P.Bull (eds.)
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
31
Conversation: an Interdisciplinary Perspective. Clevedon, U.K.: Multilingual Matters Ltd.
Romaine, S. 1999. Communicating Gender. Mahwah, N.J.: Lawrence Erlbaum.
Sacks, H., E. Schegloff and G. Jefferson 1974. A simplest systematics for the organization of turn-taking
for conversation. Language 50: 696-735.
Schegloff, E. 1982. Discourse as interactional achievement: Some uses of uh huhand other things
that come between sentences. In D. Tannen (Ed.), Analyzing discourse, text, and talk. Washington,
DC: Georgetown University Press, 71–93.
Schegloff, E. 1993. Reflections on quantification in the study of conversation. Research on Language
and Social Interaction 26 (1), 99-128.
Schiffrin , D. 1987. Discourse Markers. Cambridge: CUP.
Stenström, A. 1987. Carry-on signals in English conversation. Costerus 59: 87–119.
Stenström, A. 1994. An Introduction to Spoken Interaction. UK: Longman.
Stivers, T. 2004. ‘No no no and Other Types of Multiple Sayings in Social Interaction. Human
Communication Research 30 (2), 260–293.
Stivers, T. (2008). Stance,alignment, and affiliation during storytelling: When nodding is a token of
affiliation. Research on Language and Social Interaction, 41(1), 31-57.
Strodtbeck, F.L. and R.D. Mann 1956. Sex role differentiation in jury deliberations. Sociometry 19: 3-11.
Stubbe, M. 1978. Sex roles in conversation: a study of small group interaction. Unpublished terms paper.
Wellington: Victoria University.
Stubbe, M. 1991. Talking at cross-purposes? The effect of gender on New Zealand schoolchildren's
interaction strategies in pair discussion. Unpublished MA Thesis: Victoria University of Wellington.
Stubbe, M. 1994. What's the Score? Qualitative Analysis in Gender Research. International Journal
of Applied Linguistics, 4, 1: 3-18.
Stubbe, M. 1996. Active listening in conversation: gender and the use of verbal feedback. Australian
Linguistics Society Conference, Canberra July 1996.
Stubbe, M. 1998. ‘Are you listening?Cultural influences on the use of supportive verbal feedback in
conversation. Journal of Pragmatics 29, 257-289.
Stubbe, M. 1998. Striking a Balance: language, gender and professional identity. Engendering
Communication. Proceedings of the Fifth Berkeley Women and Language Conference 1998.
Berkeley: University of California Berkeley Women and Language Group, 1998, pp. 545-556.
Stubbe, M. 1999. Research report: Māori and Pakeha use of selected pragmatic devices in a sample of
New Zealand English. Te Reo, Journal of the Linguistic Society of New Zealand, 42.
Stubbe, M. 2003. Taking language and gender research beyond the academy. Workshop. Perception
and Realization in Language and Gender Research: An International Conference.
COSWL/IGALA, Michigan State University, East Lansing, Michigan, July 18-20, 2003.
http://www.linguistics.ucsb.edu/faculty/bucholtz/conference/stubbe.html
Stubbe, M. and J. Holmes 1995. `You know', `eh' and other exasperating expressions: an analysis of
social and stylistic variation in the use of pragmatic devices in a sample of New Zealand English.
Language and Communication, 15: 1: 63-88.
Stubbe, M. and J. Holmes 2000. Talking Māori or Pakeha: signalling identity in discourse. In Allan
Bell & Koenraad Kuiper (eds.) New Zealand English. Wellington: Victoria University
Press/Benjamins 249-278.
Stubbe, M., C. Lane, J.Hilder, E.Vine, B. Vine, J. Holmes, M. Marra, A.Weatherall 2003. Multiple
Discourse Analyses of a Workplace Interaction. Discourse Studies 5(3), 351- 388.
Stubbe, M., J. Holmes, B. Vine, M. Marra 2001. Forget Mars and Venus, let's get back to earth:
Challenging gender stereotypes in the workplace. In J. Holmes (ed.) Gendered Speech in Social
Context: Perspectives from Gown & Town. Wellington: Victoria University Press.
Talbot, M. 1998. Language and Gender: an Introduction. Oxford: Polity Press.
Talbot, M. 2003. Gender stereotypes: reproduction and challenge. In J. Holmes and M. Meyerhoff
(eds.) Handbook of Language and Gender. Blackwell: 447-467.
Tannen, D. 1984. Conversational style- analysing talk among friends. Norwood, New Jersey: Ablex.
Tannen, D. 1990. You Just Don't Understand. Women and Men in Conversation. Australia: Random
House.
CORRECTED PRE-PROOF: Stubbe, Maria. (2013). Active listening in conversation: gender and the use of verbal feedback. In
S.Yamazaki and R. Sigley (Eds). Approaching language variation through corpora: A festschrift in honour of Toshio Saito. Bern,
Switzerland: Peter Lang, pp. 367-416
32
Tannen,D. 1993. Gender and Conversational Interaction. Oxford: Oxford University Press.
Ward, N. and W. Tsukahara 2000. Prosodic features which cue back-channel responses in English
and Japanese. Journal of Pragmatics 32 (2000) 1177-1207.
Ward, N. and W. Tsukahara 2003 A study of responsiveness in spoken dialog. International Journal
of Human-Computer Studies Volume 59, Issue 5, November 2003, Pages 603-630
Weatherall, A. 2002. Gender, Language and Discourse. Women and Psychology Series. East Sussex and
New York: Routledge.
Wertheim, S., A. C. Bailee and M. Corston-Oliver 1998 (eds.). Engendering Communication.
Proceedings of the Fifth BerkelryWomen and Language Conference. April 24-26,1998. Berkeley:
Berkeley Women and Language Group.
West, C. 2006. Coordinating closings in primary care visits: producing continuity of care. In J. Heritage
and D.W. Maynard (Eds) Communication in Medical Care. Interaction between Primary Care
Physicians and Patients. Studies in Interactional Sociolinguistics 20. Cambridge: CUP, 379-415.
West, C. and A. Garcia 1988, Conversational Shift Work: A Study of Topical Transitions between
Women and Men. Social Problems, Vol. 35, No. 5 (Dec., 1988), pp. 551-575
West, C. and D. Zimmerman 1983. Small insults: a study of interruptions in cross-sex conversations
between unacquainted persons. In B. Thorne, C. Kramarae and N. Henley (eds) Language, Gender
and Society. 103-118.
White, G.M. and K.A. Watson-Gegeo, 1990. Disentangling discourse. In: K.A. Watson-Gegeo and
G.M. White, eds., Disentangling. Conflict Discourse in Pacific Societies, 3-52. Stanford, CA:
Stanford University Press.
White, S. 1989. ‘Backchannels across cultures: A study of Americans and Japanese.Language in
Society 18(1): 59-76.
Wodak, R. 1997 (ed). Gender and Discourse. London: Sage.
Xudong, D 2008. The use of listener responses in Mandarin Chinese and Australian English
conversations. Pragmatics 18:2.303-328
Zimmerman, D.H. and C. West 1975. Sex roles, interruptions and silences in conversation. In B. Thorne
and N. Henley (eds) Language and Sex: Difference and Dominance. Rowley, Massachussets:
Newbury House. 105-29.
... The fifteen included studies were conducted in the United States (n = 6), New Zealand (n = 3), Canada (n = 2), Australia (n = 1), Austria (n = 1), Sweden (n = 1), and the United Kingdom (n = 1). Thirteen studies were published in journal articles, one study was published as part of a dissertation (Pfiester, 2009), and one study was published in a book chapter (Stubbe, 2013). Study designs were either mainly quantitative (n = 13) or mainly qualitative (n = 2; Holmes, 2005;Reznik, 2004). ...
... Two studies were conducted in a laboratory setting, testing the dominance approach and dynamic approach (Ashenfelter et al., 2009;Bortfeld et al., 2001), and three studies were conducted in an institutionalized setting (Menz & Al-Roubaie, 2008;Sleath & Rubin, 2002;Waara & Shaw, 2006). The third section, on Gender Construction, describes a study combining qualitative and quantitative methods to analyze gender construction in private casual conversations (Stubbe, 2013), and a study focusing on the construction of gender in straight, trans and queer women and men in the setting of a sociolinguistic interview (Hazenberg, 2016). Additionally, two qualitative studies analysing the construction of gender in workplace interactions are also discussed in this section (Holmes, 2005;Reznik, 2004). ...
... The qualitative analysis showed that these five strategies were not exclusively associated with women or men, and demonstrated how doing mentoring or leadership and doing gender can effec-Gender Construction. Four studies focused on the construction of speaker's sex/gender through interaction in sociolinguistic interviews (Hazenberg, 2016), private casual conversations (Stubbe, 2013), or work-related conversations (Holmes, 2005;Reznik, 2004). In these studies, gender is considered to be fluid and changeable within various contexts with divergent interactional purposes. ...
Article
Full-text available
Although the question of whether women and men speak differently is a topic of hot debate, an overview of the extent towhich empirical studies provide robust support for a relationship between sex/gender and language is lacking. Therefore, the aim of the current scoping review is to synthesize recent studies from various theoretical perspectives on the relationship between sex/gender and language use in spoken face-to-face dyadic interactions. Fifteen empirical studies were systematically selected for review, and were discussed according to four different theoretical perspectives and associated methodologies. More than thirty relevant linguistic variables were identified (e.g., interruptions and intensifiers). Overall, few robust differences between women and men in the use of linguistic variables were observed across contexts, although women seem to be more engaged in supportive turn-taking than men. Importantly, gender identity salience, institutionalized roles, and social and contextual factors such as interactional setting or conversational goal seem to play a key role in the relationship between speaker’s sex/gender and language used in spoken interaction.
... In general, written popular science is an interactive genre, and when it comes to the analysis of female language, multiple studies (e.g. Hannah & Murachver, 2007;Holmes, 2005;Mulac et al., 2001;Pfiester, 2009;Rey, 2013;Roberts & Utych, 2020;Stubbe, 2013;Thompson et al., 2001) suggest various manifestations of interactivity as features of female language. ...
Article
Full-text available
Fewer women than men are involved in STEM (Science, Technology, Engineering, and Mathematics) fields and one of the reasons might be a lack of exposure to familiar linguistic structures in reading materials designed to spark interest and to recruit new scientists. Popular science, as a multifaceted genre, creates an accessible kind of exposure to the scientific world that can potentially lead to a career in that field. Considering this potential of the genre and keeping in mind that most popular science is written by men, the present study examined fifteen articles (five written by men, five by women, and five co-written by authors of both genders) for the presence of female language. Female language was differentiated from gendered language (which is tied to female stereotypes) and identified based on the frequency of certain linguistic markers found in texts produced by women. The results demonstrate that women and teams of authors that include women write differently than men do, thus confirming previous investigations into differences in linguistic production between men and women. The study suggests that exposing young women considering a career in science to popular science written by females might be beneficial in cementing their aspirations, as language has a strong connection to self-perceived gender identity.
... However, much of the "gendered" talk in media seemingly reflects heteronormative discourse, such as male-coded characters enacting hegemonic masculinity, or female-coded characters using language to index heteronormative ideals of femininity (Raymond, 2013). Yet, in reality, conversational strategies and talk that are indexical of identity markers (including gender) vary in relation to different speaker populations and the purpose of talk in a given setting (Stubbe, 2013;Yoong, 2018). Hence, by exposing children to such decontextualized representations of dialogue assigned to male-coded and female-coded characters, media appear to be promoting a form of discursive practice that homogenizes a wide range of human identities and roles (Raymond, 2013). ...
Article
Full-text available
Movies can implicitly promote social and ideological norms on a mass scale, making them powerful socialization agents, especially among children. However, Hollywood movies are no longer confined to the influence of American ideals, as media companies now have to consider the growing influence of markets such as China. With this in mind, we explore the portrayal of gender, power, and gendered roles across two versions of Disney’s Mulan (1998 and 2020). Specifically, we explore male-coded and female-coded characters’ talk for portrayals of gender and the enactment of assigned roles through conversational strategies and the content of talk. Findings indicate a subtle shift in the distribution of “dominant” discourse between the two versions, despite female-coded characters being framed as dutiful wives, brides-to-be, and/or mothers in both movies. Specifically, in Mulan 2020, male-coded characters are portrayed as more “feminine” through their talk, while female-coded characters—particularly Mulan and Xianniang—are portrayed as more “masculine”, further highlighting a recent trend for more nuanced portrayals of gender in Disney movies. Based on these results and others, we argue that the portrayal of gender in Mulan 2020, while still primarily associated with heteronormative roles in service of a patriarchal world, has undergone subtle changes that may reflect American and Chinese influences.
Article
Full-text available
Introduction: Backchannels, i.e., short interjections by an interlocutor to indicate attention, understanding or agreement regarding utterances by another conversation participant, are fundamental in human-human interaction. Lack of backchannels or if they have unexpected timing or formulation may influence the conversation negatively, as misinterpretations regarding attention, understanding or agreement may occur. However, several studies over the years have shown that there may be cultural differences in how backchannels are provided and perceived and that these differences may affect intercultural conversations. Culturally aware robots must hence be endowed with the capability to detect and adapt to the way these conversational markers are used across different cultures. Traditionally, culture has been defined in terms of nationality, but this is more and more considered to be a stereotypic simplification. We therefore investigate several socio-cultural factors, such as the participants’ gender, age, first language, extroversion and familiarity with robots, that may be relevant for the perception of backchannels. Methods: We first cover existing research on cultural influence on backchannel formulation and perception in human-human interaction and on backchannel implementation in Human-Robot Interaction. We then present an experiment on second language spoken practice, in which we investigate how backchannels from the social robot Furhat influence interaction (investigated through speaking time ratios and ethnomethodology and multimodal conversation analysis) and impression of the robot (measured by post-session ratings). The experiment, made in a triad word game setting, is focused on if activity-adaptive robot backchannels may redistribute the participants’ speaking time ratio, and/or if the participants’ assessment of the robot is influenced by the backchannel strategy. The goal is to explore how robot backchannels should be adapted to different language learners to encourage their participation while being perceived as socio-culturally appropriate. Results: We find that a strategy that displays more backchannels towards a less active speaker may substantially decrease the difference in speaking time between the two speakers, that different socio-cultural groups respond differently to the robot’s backchannel strategy and that they also perceive the robot differently after the session. Discussion: We conclude that the robot may need different backchanneling strategies towards speakers from different socio-cultural groups in order to encourage them to speak and have a positive perception of the robot.
Chapter
Conversational turn-taking
Chapter
This chapter explains what is meant by the term "language ideologies" and goes on to examine gender differences in language use as one area in which ideological representations are common. After pointing out that these representations can differ significantly across cultures and through time, it discusses the emergence in contemporary Western societies of a specific representation of male-female difference, here labeled "the myth of Mars and Venus," in popular self-help literature, and the subsequent rise of new representations which place particular emphasis on biological differences. It concludes by considering the relationship between language ideologies and real-world linguistic practices, and suggests that this relationship makes the study of language ideologies an important part of the study of language and gender more broadly.