Tuesday, 30 October 2012

Do you smile with your nose?

Do smileys ever have a nose?

Emoticons are a good way of representing what gets lost when we switch from speaking to texting. We can’t use intonation or facial expressions to show whether we’re joking or are sad, so we use an emoticon to do the job. The emoticons people use can vary, though, as Tyler Schnoebelen discovered when he analysed nearly 4 million American English tweets that included at least one of the most frequently used emoticons. Just some of the variations that people used are shown in the Table.

Number in the corpus
Percentage of all emoticons in the corpus
: )
; )
: (
big smile
: D
: -)
; -)
     70, 618
: -(

Counts and percentages of emoticons in the American English Twitter corpus analysed by Schnoebelen

Schnoebelen showed that the variants corresponded to different types of users, tweeting with different vocabularies. His statistical analyses revealed that the most pervasive distinction was between emoticons with noses and those without noses. He therefore set out to discover whether emoticons with and without a nose, such as :) and :-) , mean the same thing. He did this by looking at how they patterned with other aspects of tweets.
Tweets cannot be longer than 140 characters, so you might expect people who send longer tweets to use emoticons without a nose, to save a character. But it turned out that people who used noses wrote longer tweets, not shorter ones. They also avoided abbreviations like thru. These ‘nosers’ made few typos and spelt words correctly. Overall, then, their language could be described as more standard. The ‘non-nosers’, by contrast, seemed to want to be more non-standard.  They tended to mis-spell words such as ‘tomorrow’ as tommorow, they dropped the apostrophe in contractions such as wasn’t, and they used more taboo words and more expressively lengthened words (like soooo or yummm). They also used more emoticons overall. They seemed to be younger than the ‘nosers’, keeping up with a younger set of celebrities and sending them positive vibes.

Schnoebelen explains that emoticons with noses were the first to be used, so for a while they were the historically ‘standard’ forms. This meant that people who were interested in presenting themselves as nonstandard had to change them, and remove the noses.

So, do emoticons with noses mean the same as those without a nose? Schnoebelen reminds us that meaning is an emergent property of social relations, not something that a symbol has in itself. He gives as an example a bouquet of roses, which is meaningful because there are lovers, patients, doctors and florists to give it meaning: the interpretation is shared by people we’re familiar with, using familiar interpretive schemes. To understand the meaning of emoticons, then, we need to think not only about the emotions they can convey but also who uses them and when. A smiley can tell us how the person feels about what they are tweeting, but it also tells us something about the kind of relationship they want to establish with the people they are tweeting.
Schnoebelen, Tyler (2012) Do You Smile with Your Nose? Stylistic Variation in Twitter Emoticons. University of Pennsylvania Working Papers in Linguistics 18 (2): 117-125

This summary was written by Jenny Cheshire

Friday, 26 October 2012

Candidate understandings in conversation

                        A: I just bumped into whatsisname….the guy in the                                 next office…in [the canteen]
                        B:                     [you mean] John?
                        A: yes, that’s right

In the above imaginary conversation, the listener (B) has interrupted the speaker (A) to help out with a name that the speaker cannot readily supply. In this instance, the interruption to the flow of the speaker’s talk can be seen as helpful (or ‘affiliative’); in other words, the listener’s interruption is momentary and is heard as trying to check or clarify an aspect of what the speaker means to say. Among linguists, these interruptions are referred to as candidate understandings.

Researcher Charles Antaki has been taking a closer look at a collection of candidate understandings and suggests that they may not always be designed to be so co-operative. He argues that they can sometimes signal dissatisfaction with the speaker’s talk and that they can be a serious hindrance (or be ‘disaffiliative’) to the progress of continuing that talk. Antaki provides the following example from a conversation between Dana and Gordon:

                        G: I managed to get home in time for my music                                      lesson at five…thirty?...h[hhhh
                        D:                                   [Mm hm?=
                        G:=hu – uh dashing back at (.) at a grand sixty miles                             ‘n hour in..Malcolm’s car it nearly shook itself to                                 pieces..hhhh he wz zipping round the roads:- (0.3)
                        D: for your music lesson
                        G: yeah that’s right, .hhhhh
In the first example above, the listener offers some fresh information (the name of the person) to try and meet a need in the speaker’s utterance. However, in the second example, there is no obvious problem with G’s story and D does not offer any missing information. Instead, D’s utterance of ‘for your music lesson’ is merely a repeat of something that G has already said and has the effect of pulling G up short in the telling of his story. Antaki proposes that this kind of candidate understanding points to something objectionable about the speaker’s utterance and requires the speaker to do something about it. In this case, D appears to be objecting to G’s failure to stick to (what D considers to be) the significant point of the story and it has the effect of bringing the story to an abrupt end. The two speakers in the examples also seem to recognize the different functions of the candidate understandings. The speaker in the first example responds immediately with ‘yes, that’s right’ but the speaker in the second example stops the narrative of the journey home, audibly hesitates and then answers ‘yeah that’s right’.

Antaki argues that in the second type of candidate understanding there is an underlying notion that the listener has spotted something that he or she ‘knows better’ than the speaker and is requiring the speaker to do something about it. Clearly, then, in the above example D feels that she has the right to exert her perspective on the story and alter the direction of what is being said. Of course, it seems highly likely that the relationship between the interlocutors, the setting in which the interaction takes place and perhaps the type of interaction will all have a role to play. Furthermore, as the author points out, these ‘unhelpful’ understandings can also be used for ironic effect. Discounting ironic uses, though, Antaki points out that sometimes at least, the practice shows that ‘ telling someone that you know better is equivalent to telling them what to do’.
Antaki, C. (2012). Affiliative and disaffiliative candidate understandings. Discourse Studies 14(5): 531-547.

doi: 10.1177/1461445612454074

This summary was written by Sue Fox

Tuesday, 23 October 2012

Mumbling is macho

Men who pronounce their words less distinctly are perceived as more macho, or so the research of Kevin Heffernan suggests. Many previous studies have found a gender difference in how we articulate our words, with men tending to articulate less precisely than women. However Heffernan is the first linguist to show that some men articulate more precisely than other men, and that this appears to cause listeners to perceive them differently.

Heffernan recorded 8 radio DJs from XM Satellite Radio, a North American for-profit company that broadcasts over 170 stations. Between them the DJs covered 6 musical genres: classical, Country and Western, popular, hard rock, punk and heavy metal. He then asked university students (9 men and 6 women) to judge the DJs’ voices on 10 different attributes, such as how friendly they sounded or how technically competent the students thought they were. Statistical analyses were used to group attributes that received similar scores into 4 groups that Heffernan interpreted as indicating the DJs’ personality, regional accent, social class and masculinity. Masculinity correlated with the scores for machismo, self-confidence, composure and estimates of the DJs’ physical size.

Heffernan then measured four features of the DJs’ pronunciation, to see whether there was any correlation between precise articulation and the social attributes that the DJs had been judged to possess. One feature was how distinct from each other were pronunciations of the six vowels /i, e, æ, u, o, ɑ/, while two others concerned types of contrast in the length of different vowels: for example, the vowel in the word beat is usually longer than the vowel in bit, though the actual difference in length varies from one speaker to another. The fourth feature analysed was the pronunciation of the final consonant in the word with. In fast speech this sometimes sounds like a /d/ or a /t/, depending on the type of sound that follows. On each of these four measures, Heffernan found a significant correlation between the DJs’ articulation and the ratings they had been given for their perceived masculinity: those DJs judged as sounding more masculine produced phonetically less distinct speech.

There was little or no relation between the judgements of the DJs’ personality or social class and the precision of their articulation. Interestingly, though, judgements of their regional accent also correlated to some extent with the clarity of their articulation. Perhaps this is because regional accent is sometimes associated with masculinity: Heffernan points out that Southern US accents, for example, have been associated in experiments with male-dominated institutions such as the American military.

Heffernan does not comment on the music that the DJs were associated with but, for the record, the DJ who received the highest scores on the masculinity group of attributes played heavy metal while the DJ with the lowest scores played classical music. However the conclusion of the paper relates not to the music but to the fact that precision in articulation can send subtle messages about the personal characteristics of the person who is speaking – in this case, about how macho the speaker is.

Kevin Heffernan (2010) Mumbling is macho: Phonetic distinctiveness in the speech of American Radio DJs. American Speech 85 (1): 67-90.

doi 10.1215/00031283-2010-003

This summary was written by Jenny Cheshire

Tuesday, 16 October 2012

Agreement in the House of Lords? Not likely!

A vintage engraving from 1861 showing tumult in the House of Lords 

Anyone who has watched Question Time in either the House of Commons or the House of Lords of the British Parliament will know that there is a certain kind of institutional rudeness which is acceptable as part of the ritualized confrontational encounters that take place there. It is the job of MPs and Members of the House of Lords to question government ministers on matters for which they are responsible. Disagreement is therefore part and parcel of this particular context and interaction.

However, the participants are also bound by the rules of the respective Houses and perhaps this is especially so in the House of Lords, where some of the highly formal traditions have been in place since the 14th century and which account for the sometimes archaic language and behaviour. So, in this situation, how do the Lords ‘do’ disagreement? Researcher Jessica Robles collected around 300 video recordings and transcripts of the meetings (from the official document called Hansard) of interactions in the House of Lords, available from the United Kingdom Parliament website. 

In this paper she provides an in-depth analysis of two examples of disagreement from the oral question sessions of the House of Lords (you can read the whole article here) to support her main argument, which is that disagreement is achieved through what she calls ‘talking around the issue’. Robles argues that there are three dimensions to the practice of talking around the issue. These are: 1) institutional positioning, in which the talk itself positions the participants within their current role and within the institutional context of the government structure as well as in opposition to others. 2) display of emotionality, a way of talking around the issue by appearing to disagree but not actually specifying what the disagreement is about. The talk contains markers of ‘feelings’ that cue people to a particular attitude being presented – in one example, Robles includes such things as lack of fluent speech, use of emphatics such as ‘absolutely’ and ‘precisely’ and nonverbal expressions such as frowning, rolling the eyes and shaking the head. 3) orientation to the issue , in which talking about the focal point of an argument may be avoided explicitly but is referred to implicitly. By using these three practices, disagreement is thus carried out through indirect orientations to an implied issue.

Overall, these practices sound very familiar, not only in the House of Lords but also perhaps in political debates more generally, where speakers talk around an issue by addressing the issue through indirect strategies. These strategies appear to be shaped by the context in which they take place but Robles argues that they are not bound to this context. While talking around the issue may be considered as ‘beating around the bush’ in everyday interactions, she suggests that the strategies could be investigated as a way of dealing with disagreement in other relevant contexts. 
Robles, Jessica. (2011). Doing disagreement in the House of Lords: ‘Talking around the issue’ as a context-appropriate argumentative strategy. Discourse and Communication 5(2) 147-168.

doi: 10.1177/1750481310395452

This summary was written by Sue Fox

Friday, 12 October 2012

Workshop for English Language teachers

Analysing spoken English: Resources and techniques for teachers

Workshop for English Language teachers to be held at Newcastle University, UK, on Saturday, 8 December 2012.

The workshop is hosted by the School of English Literature, Language and Linguistics and is organized by staff at Newcastle University, with contributions from colleagues at Queen Mary, University of London, and the University of Ottawa, Canada.

Here’s what other teachers have said about the workshops:

“fantastic value for money”

“excellent resources”

“these workshops need to be advertised more widely. Many other English teachers would come if they knew about the quality – and the price!”

programme outline
9:30 – 10:00: Arrival and registration

Part 1: Discourse-pragmatic features: What are they? How do they develop? Why do we need them?  
10:00 – 10:15: Introduction: What are discourse-pragmatic features?  
10:15 – 10:45: The usage, function and spread of innit?  
10:45 – 11:15: Reporting speech: The use of quotatives in spoken language  
11:15 – 11:45: Linguistic irritants or indications of communicative competence?: Children’s use of general extenders
11:45 – 12:00: Q-A-session  

12:00 – 12:45: Lunch break

Part 2: English Language teaching: On-line databanks and resources
12:45 – 13:15: Teaching the google generation
13:15 – 13:45: Linguistics Research Digest  
13:45 – 14:00: Q-A-session  

14:00: Workshop close

To register your interest in the workshop please e-mail Melanie Birch (melanie.birch@ncl.ac.uk) who will then raise an invoice for you. Please include in your e-mail your own name, the name and full postal address of your school/institution, a contact phone number, your e-mail address and, where this is different from your own details, the name and contact details of the person you would like us to invoice for the event.

The registration fee is £48 and covers administrative costs, the costs for production of the resource booklet and a buffet-style lunch. The deadline for registration and receipt of payment is Thursday, 15 November 2012. Registration fees are non-refundable after this date, except in the unlikely event of the workshop being cancelled.

Details about the exact venue, travel etc. will be circulated in due course. However, if you have any queries in the meantime, please contact the local organiser, Dr Heike Pichler, heike.pichler@ncl.ac.uk

Tuesday, 9 October 2012

The complexity of conversations

We may not realise it but conversations are governed by rules

Even though we may not realise it, everyday informal conversations are highly coordinated events which are managed by the participants on a turn-by-turn basis. Our role in conversations constantly alternates as we either take up the task of acting as the current speaker or the current hearer. Conversations can be analysed in terms of turn-constructional-units (TCUs) and transition relevance places (TRPs). A TCU is meant to describe a piece of conversation that may comprise an entire turn and a TRP is a point in the conversation where the talk could legitimately pass from one speaker to another. In reality, a hearer doesn’t always take up the opportunity to speak at every TRP and so the same speaker will often continue their turn with a new TCU. This means that one person’s whole turn in a conversation may comprise of multi-TCUs before they complete their turn and another participant takes up the speaking role.

Emanuel Schegloff, a pioneer in the field of Conversation Analysis, has been looking at multi-TCU turns and/or multi-turn sequences in which ‘answering’ is being done. Typically, these may be sequences in which a question is posed by one participant and which is then ‘answered’ by another participant. In the process of answering, several TRPs may occur but if the hearer doesn’t recognize them as such or chooses not to take up the talking role, then the same speaker will often continue speaking. So how can speakers end these sequences and convey to the hearer that they have nothing further to say on the matter? Schegloff argues that one practice for showing that a particular TCU of a multi-unit turn is meant to be its last is to repeat an element from the start of the turn and/or the sequence. He provides several data extracts which support his argument. 

In the following example, ‘Joey’, who lives with his father ‘Tony’, has driven to visit his mother ‘Marsha’ (who lives in another part of the state of California) for the weekend. Tony has rung Marsha to find out when Joey will be returning home.

     Tony:      Well I wz wondering when ‘e left
      Marsha:   ‘hhh Uh:(d) did Oh: .h Yer not in on what ha:ppen’.(hh)(d)

Marsha then ‘answers’ Tony’s indirect question by explaining that Joey is travelling home by plane because his car had been broken into and he couldn’t drive it home. At several points in the telling there are possible TRPs but Tony does not take them up; he either remains silent or uses a response such as ‘uh huh’ which acts as a signal that he expects Marsha to continue talking. Finally, Marsha shows that she is ending her turn by saying:

     Marsha:   En I left him there (at the airport) et abou:t noo:n

Interestingly, the two uses of ‘left’ are very different. In the first use, the ‘left’ has the meaning of ‘departing’ and the one doing it is Joey. In the second use, the ‘left’ has the meaning of ‘abandoning’ and the one doing it is Marsha. Note also that the two uses of ‘left’ are by the two different speakers. Nevertheless, Schegloff argues that there is sufficient evidence to show that this repetition of an element from the start of the turn/sequence is enough to serve as a way of marking that the turn is complete.

Schegloff goes further and proposes that the practice of moving to close by returning to an element of the start also extends to closing sequences more generally and not only to second parts of question/answer sequences and that ‘returning to the beginning’ can occur over substantially long stretches of conversation.
Schegloff, Emanuel A. (2011). Word repeats as unit ends. Discourse Studies 13(3): 367-380.

doi: 10.1177/1461445611402749

This summary was written by Sue Fox 

Friday, 5 October 2012

Uh huh, no way!

reactions to reactions can change the way conversations develop

A lot of the focus in conversation analysis relates to how speakers construct and manage their turns in conversation and how listeners react to what is being said.  However, as Neal Norrick points out, there has been little attention to how a speaker reacts to their listener’s response. For example, consider the different responses of Speaker A to the following feedback from Speaker B below:

1)         A: I had lectures going on till 7 at night
            B: mhm
            A: and so I wouldn’t get home till really late

2)         A: I had lectures going on till 7 at night
            B: no way
            A: it’s true, something went really wrong with the timetable                     that year

We can see that while the mhm response in (1) doesn’t interrupt speaker A and allows them to carry on with their story, the response of no way in example (2) instead causes speaker A to defend their initial statement and not continue with their story about getting home late. Therefore, these types of listener responses function in different ways with respect to how much of a reaction they get from the speaker.

Norrick describes different feedback tokens in terms of three categories: continuers, assessments and information state tokens.  Continuers are those contributions such as mhm and uh huh which show the speaker that the listener is listening and that the speaker can carry on their turn without being fully interrupted. Continuers are also the least likely to elicit any kind of response from the speaker. Assessments, instead, are tokens such as wow, gosh or yuck and are more likely to prompt a response from the speaker.  They provide some kind of emotional response or evaluation to what has been said.  Finally, the most likely responses to get a reaction from a speaker are information state tokens. These include so, really, oh and yeah which can be used to negatively challenge what the speaker has said (e.g. by showing scepticism or sarcasm) and require the speaker to deal with them in their subsequent turn (such as example 2 above).

Norrick suggests that these listener responses can form a continuum of intrusiveness, depending on how noticeable they are to the speaker and how much they need to be addressed in the discourse. However, they can also be judged with respect to politeness.  For example, the production of mhm and uh huh show consideration and attentiveness on the part of the listener. If they are not produced, this can indicate lack of interest or attention and cause the speaker to abandon their turn.  Similarly, assessments mark positive politeness in that they can signal appreciation and involvement in the discourse.  In contrast though, information state tokens (such as well or so what) can be viewed as the least polite forms as they challenge the speaker’s motivation for saying what they do.

With these points in mind, Norrick cautions that a rigid scale relating to consideration and politeness would be unwise as we need to take into account various social aspects, such as the relationship between the interlocutors. However, he also asserts that more research should be done cross-linguistically so that we can gain a wider appreciation of conversational features across cultures.
Norrick, Neil  2012. Listening practice in English conversation: The responses responses get. Journal of Pragmatics (44:566-576)

doi: 10.1016/j.pragma.2011.08.007

This summary was written by Jenny Amos

Tuesday, 2 October 2012

Persuasion and ideology in Cosmetic Surgery advertising

In pursuit of the ideal body?

There is no doubt that cosmetic surgery in the UK and the US (and many other countries) is big business and is still a growing industry. Over 10,000 women in the UK underwent surgery for breast enlargement in 2011, and one can only hazard a guess at the number of people undergoing non-surgical procedures, such as botox injections. Research studies have shown that media representations of cosmetic surgery tend to present operations as a quick and easy solution to achieving a changed (and ideal) body and, in the process, feeling better about one’s self, without paying very much attention to the risk or recovery times involved. So how do cosmetic surgery advertisements convey these ideas?

Researchers Maria Martinez Lirola and Jan Chovanec collected a set of 20 cosmetic surgery leaflets handed out to passers-by in the streets of Alicante (Spain) by a leading Spanish cosmetic surgery clinic during the years 2007 to 2009. Their research aimed to show how the advertisements combined language and visuals to persuade people, mainly women, to have cosmetic surgery. They analyse two of the leaflets in depth but they also make relevant points about the leaflets in general.

The visual characteristics involved in the construction of meaning within the leaflets are described in terms of dominance, frames and information. It was noted that in 19 out of the 20 texts the dominant picture was of a young, white woman, usually scantily clothed in a seductive pose and accompanied by a man. The researchers state that the woman is the main focus because although cosmetic surgery is offered to men, women are the primary clients. All of the men and women in the pictures were in fact white and young, representative of the main intended audience. The use of frames refers to other visual elements of the text and where they are placed. In the leaflets, pictures of other people are much smaller than the main picture and placed in frames around the side. The layout of information is also important; new information is usually placed to the right, the top of the page is associated with the ‘ideal’ and the bottom associated with the ‘real’. The leaflets contained the logo and name of the cosmetic surgery clinic at the top, placing the perception of cosmetic surgery within a medical frame. The bottom part of the leaflets usually showed the bigger parts of the body and any display of ‘action’ (e.g. in one leaflet the woman plays with the strap on her underwear). The centre of such texts usually coincides with the main information.

The researchers note that the linguistic elements of the text enhance and complement the visual aspects of the advertising messages. The texts used ‘highly positive evaluative expressions and instructional phrases’ of the kind shared with other kinds of advertising. Adjectives such as ‘best’ and ‘highest’ were used and rhetorical questions (e.g. ‘what else can we ask for?’) were also frequently used. Common persuasive techniques running through the texts related to competitive pricing, a free gift, the rhetoric of numbers (e.g. ‘over 10,000 patients’), claims about the professionalism of staff and the positioning of cosmetic surgery within a medical frame. All of these devices are aimed at persuading potential clients that surgery is safe and affordable and, the researchers claim, the advertisements also play subtly on inner fears of ageing.

Martinez Lirola and Chovanec point out that images of ‘ideal-looking’ women serve to act as a model towards which other women could (and should) aspire. The surgically enhanced body suggests that it is the key to ‘women’s self-esteem, self-confidence and physical perfection’ and that it is ‘the target of male voyeuristic desire’. Of course, as the researchers rightly note, the persuasive techniques also fulfill the aim of generating profits for the cosmetic surgery providers.
Martínez Lirola, María and Chovanec, Jan (2012). The dream of a perfect body come true: Multimodality in cosmetic surgery advertising. Discourse and Society 23(5): 487-507.

doi: 10.1177/0957926512452970

This summary was written by Sue Fox