Master Thesis
You still say hello!
A corpus study of the greeting words hi, hey and hello in spoken American English
Author:Oskar Rydblom Supervisor: Magnus Levin Examiner: Mikko Laitinen Semester: VT2014
Subject: English Level: Advanced Course code: 5EN01E
Abstract
This study examines the usage of the three greetings words hi, hey and hello in
combination with greeting questions, such as how are you, and terms of address, such as Bill, Mr Jones or Dad, in spoken American English. The investigation identifies some of the patterns and rules that govern greetings sequences and to what extent social factors like gender, age, social class and social context inform the speaker‘s choice of greeting. Theories within the area were then used to identify greetings in the Longman Spoken American English Corpus that were then entered into a database along with information about the speakers. Using filters the greeting pairs where then sorted by different factors, such as gender, to extract quantitative data on the distribution and frequency of usage.
The results showed variation influenced by social factors and formulaic patterns of greeting sequences. The study concludes that hi is the standard and most frequent greeting in spoken American English, hey is a more vernacular form and hello seems to be more frequently used when respecting the boundaries of the addressee. Greeting sequences also follow a predictable pattern based on the greeting word used or the greeting questioned asked. More formal terms of address (for example Mr Lastname) are used in more formal settings or when there is a significant age difference between speakers. If a greeting question (how are you?) is posed it is usually returned, answered or both, and if answered often with a short reply providing minimal information (fine).
Hence, both the choice of greeting word and the use of terms of address and greeting questions are influenced by social factors like age, gender, social class and formality of setting according to observable patterns.
Keywords: Hey, hi, hello, greetings, adjacency pairs, politeness, Longman Spoken American English Corpus, corpus linguistics sociolinguistics, phraseology
Table of Contents
1 Introduction 1
2
Theoretical background 32.1 Discourse 3
2.1.1 Analysing discourse 4
2.1.2 The formulaic nature of discourse 6
2.1.3 The social aspect 8
2.2 Politeness 10
2.2.1 Politeness, class and social groupings 13
2.2.2 Politeness, age and gender 15
2.2.3 Politeness and culture 16
2.3 Greeting words and greeting questions 17
2.3.1 Definitions of words and terms used in this study 21
3 Material and Method 23
3.1 The Longman Corpus of Spoken American English 23
3.2 The pilot study 26
3.3 Creating a database 29
3.4 Method of analysis 38
3.5 Problems and limitations 39
4 Results 41
4.1 Frequency of usage 41
4.1.1 Variation according to age 43
4.1.2 Variation according to gender 47
4.1.3 Variation according to social class and context 50
4.2 Adjacency pair relationships 55
4.2.1 Greeting questions and terms of address 60
4.2.2 Gender, age and social context in greeting sequences 65
5 Discussion 72
6 Conclusions 75
List of references 77
1
1 Introduction
The importance of language in social interactions has long since been known and books on etiquette have been quite popular throughout the ages, one of the most famous ones being Erasmus‘ On the Civility of Children's Conduct published around 1530. Knowing how to successfully navigate social situations has therefore long been considered very important. The idea that the use of language could be viewed as performing an action, put forward by Austin (1962), Searle (1969) and others, resulted in a strong interest in the study of what is now referred to discourse analysis. The maxims formulated by Grice (1975) suggested that interactions operate within a framework, obeying rules that can be studied and expressed scientifically. These rules for social interaction and interactions between cultures drew the attention of many linguists, anthropologists, philosophers and sociologists. The studies produced at the time often centred round the concept of politeness. The notion of face, presented by Goffman (1967) and further explained in section 2.1.3, played an important role in this field. The focus of this paper is on one of the more basic features of social interaction, the greeting.
Greeting words and greeting questions are a universally occurring phenomenon.
While they may vary between cultures they tend to occur in one form or another in the civilizations we know. They are a vital part of social interaction and among the first form of social interaction that children acquire (Greif & Gleason 1980). Greetings are the traditional way of initiating a spoken interaction between two or more people and it is often considered impolite to leave them out (Labov & Waletsky 1967). It is therefore important for speakers to be aware of these social rituals so that interactions function smoothly. Since greetings are often formulaic in nature, the area of linguistics known as phraseology is also one of interest.
This study sets out to examine how the usage of, and response to, the greeting words hi, hey and hello varies in different social situations and what patterns these greeting words, sometimes combined with terms address or greeting questions, follow.
The example below shows a greeting initiated using the greeting word hi and a term of address (a name) and a response to the greeting that contains the greeting word hi and a greeting question.
<?> Hi Alex.
<0932> Hi how are you? (LSAC)
The paper aims to fill a gap in research since previous studies focus mainly on the general function of greetings, or more narrow studies of a particular group or
2 community, and not the usage and variety of choices made by English native speakers in different social groups. The study conducted in this paper provides empirical data to support these more general theories. The first section provides a summary and
discussion of the progress made in discourse, politeness and phraseology since the second half of the 20th century. Insights gained from these different fields were then used to pose a framework for dealing with greeting phrases that can be tested
empirically. Identifying flaws and benefits of this kind of model of using a corpus could prove useful to further studies on social norms in language. This study is focused on the greeting words hi, hey and hello and conducted using principally the Longman Spoken American Corpus (from now on referred to as LSAC in this paper). The aim, research questions and scope below show the focus of the investigation.
Aim:
The aim of this study is to show that greetings are formulaic in nature and that the choice of greeting words and questions are connected to social factors such as gender, age, social class and context.
Research questions:
How does the usage of the greeting words hi, hey and hello vary in regards to gender, age, social status and context in spoken American English?
What if any patterns do greeting sequences follow?
If patterns do exist, how can they be explained and how are they connected to gender, age and context?
Scope:
The study looks primarily at 366 greetings and responses in American English found in LSAC and how the greeting words hi, hey and hello are used and combined with terms of address and greeting questions. The social factors that may influence change taken into consideration are age, gender, social class and context (the situation in which the conversation takes place). These 366 pairs do not account for all the greetings in LSAC, only the ones where one of the three greeting words selected was used and sufficient information on speakers was available. Other corpora are used only for comparison of general frequency to assess how representative LSAC is.
3
2 Theoretical background
The theoretical background for this study is divided into three main sections. The first section deals with some of the main theories on discourse and discourse analysis as well as a short chapter on phraseology that expands on frequently occurring words and phrases. In the second section the field is narrowed to the study of politeness in which cultural aspects of language are also discussed. The last section deals with greeting words and phrases specifically.
2.1 Discourse
Before diving into the topic of discourse the term discourse needs to be defined, as opinions often differ on the details. In their introduction Jaworski & Coupland (1999) cite a number of different definitions of the word discourse to illustrate the broadness of the term. Two of them put together aptly sum up the main view this thesis takes on discourse:
―The analysis of discourse is, necessarily, the analysis of language in use. As such it can not be restricted to the description of linguistic forms independent of the purposes of functions which these forms are designed to serve in human affairs.‖
(Brown & Yule 1983:1)
―Discourse constitutes the social. Three dimensions of the social are distinguished – knowledge, social relations, and social identity – and these correspond respectively to three major functions of language … Discourse is shaped by relations of power, and invested with ideologies.‖
(Fairclough 1992:8 cited in Jaworski & Coupland 1999:2)
In simpler terms language cannot easily be separated from social and cultural influences or the intentions of the interlocutors since any form of interaction is always influenced by the social settings in which the interaction takes place, as well as by those who
4 participate in it. As Jaworski & Coupland (1991:49) point out discourse includes far more than just the literal meaning of the words in an utterance; it is about fulfilling goals and purposes, thus most of the information is implied (pragmatic). The problem with accepting such a definition is that accounting for all the factors that influence an interaction is practically impossible. This paper therefore agrees with Jaworski &
Coupland‘s (1999:37) view that all research is ―inherently imperfect‖ and that to
achieve any level of clarity one needs to look at many differing perspectives. As Duranti so eloquently puts it:
―Having a language is like having access to a very large canvas and to hundreds or even thousands of colors. But the canvas and the colors come from the past. They are hand- me-downs. As we learn to use them, we find out that those around us have strong ideas about what can be drawn, in which proportion, in what combinations, and for what purposes. As any artist knows, there is an ethics to drawing and coloring as well as a market that will react sometimes capriciously, but many times quite predictably to any individual attempts to place a mark in the history or representation or simply readjust the proportions of certain spaces at the margins. … Just like art-works, our linguistic products are constantly evaluated, recycled or discarded.‖
(Duranti 1997: 334 cited in Jaworski & Coupland 1999:37–38)
When dealing with variables such as influences from society, implicatures and other data of this sort, the view and perception of the researcher is an important factor. The empirical study carried out in this paper works with looking at quantative data from a corpus and therefore aspires to provide objective observation while realizing that some subjective judgements are unavoidable when explaining patterns in the data.
2.1.1 Analysing discourse
Having discussed the definition and ideologies pertaining to discourse this section moves on to models and theories on how to analyse discourse. One of the basic ideas of discourse analysis is the likening of language usage to performing different types of actions, sometimes referred to as speech acts. Austin (1962) set out to distinguish between statements (considered to be strictly descriptive or informative) and acts (such
5 as warnings). In order for these acts to be performed they have to fulfil what Austin refers to as felicity conditions, which are conditions that determine whether such an act can be used or not. The speaker needs to be sincere in his intentions and the context must be such that it allows the act to be performed. In order for someone to talk about their children they need to have children for example. So, a response to a greeting question such as How are you? can naturally only occur if the question is asked.
Grice (1975:78) takes the position that those participating in discourse are bound to follow certain rules; he summarizes this in what he calls the ―Cooperative Principle‖
or CP.
―Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.‖ (ibid)
The work and theories presented by Austin and Grice were to many the stepping stone for creating frameworks for analysing discourse discussed later in this section.
Gumperz (1977:99) also considers conversation to be ―a cooperative endeavour, subject to systematic constraints‖. He argues that often statements are made in a certain style to give associations to a specific context. A friend commenting on another‘s problems by saying And how does that make you feel? is only perceived as funny if the friend is aware that this is a phrase typically associated with psychiatrists. The tone of voice (possibly imitating Dr Phil) and posture may provide additional clues. Gumperz (1977:101) refers to these clues as ―contextualisation cues‖ and suggests that they allow speakers to ―associate styles of speaking with contextual presuppositions‖. In other words our cultural and social background plays an important part in understanding implicatures, or the pragmatic aspects of discourse. For similar reasons Schlegoff (1991: 109) considers Conversation Analysis (CA) to cross over from linguistics into fields such as sociology, anthropology and psychology. Thus, it seems apparent that any analysis of discourse needs to take cultural aspects and societal values into account when looking at a text.
These theories would suggest that an interaction, such as a greeting, requires certain felicity conditions to be fulfilled, that the participants are willing to provide relative contributions and that social and cultural factors will inform the choices and understanding of the exchange. Expressions that cross genres may also be an indication
6 of on-going change in language or that other social factors are in play that influence speakers‘ choices.
2.1.2 The formulaic nature of discourse
Many parts of interactions follow a more specific format than others. One such format described by Sacks & Schlegoff (1973:265-266) is the utterance pair or ―adjacency pair‖, two utterances connected and dependant on each other. Sacks & Schlegoff list the following criteria for an adjacency pair:
(1) It needs to be two utterances in length
(2) There needs to an ―adjacent positioning of component utterances‖, meaning that the two utterances should not be separated by other utterances
(3) The utterances need to be spoken by different speakers
(4) There is a ―relative ordering of parts‖ within the pair (one comes before the other)
(5) ―Discriminate relations‖ exist. This roughly means that the first part of the pair to some degree limits and decides the choice of the second part.
These pairs occur in many situations such as greetings, goodbyes and when thanking someone. Sacks & Schlegoff (1973: 266) argue that the reason for this is that the second part (utterance) of the pair works as receipt. It signals that the responding interlocutor has understood the first interlocutor‘s purpose and accepts it. Their study suggests that in transitions, such as changing the topic, ritualistic phrases (Things always work out for the best) and adjacency pairs (Okay? Alright.) are very common (ibid:269). The fourth condition of ―relative ordering‖ could be troublesome if adapted too strictly as Sacks &
Schlegoff seem to refer mainly to the functions of the utterances; in other words a question comes before an answer, but not the other way around. Thus, Alright? Okay would also be a possible adjacency pair; or in the case of greetings it would be possible to respond to hi using hello and to hello using hi, but not possible to respond to a
greeting before you are greeted. So in short this theory dictates that the person initiating
7 a greeting sequence will in some way limit the response options, producing a receipt from the person being greeted.
Related functions have been found in discourse markers, Shiffrin (1988:274- 284) describes how the marker oh is used to manage information. It is often used to show that the speaker is paying specific attention or want to draw attention to certain information. Such as correcting yourself or someone else (I was fifteen, oh wait.
Sixteen.), responding to unexpected information (Oh, I didn‟t know that) or to intensify a statement (Oh yes it is!). To Shiffrin (ibid:286) the marker is an indication that the speaker is attentive and wishes to show their focus on something, hoping that the other interlocutor will focus on the same thing. It shows active participation and
acknowledgement of the fact that a speaker‘s knowledge changes throughout a
conversation. Since the greeting words considered in this study are sometimes also used as discourse markers the function and workings of such particles do need to be
considered in order to identify the purpose of these words in the text.
Tannen & Wallat (1987: 348-349) argue that often the circumstances
surrounding an utterance are more important than the utterance itself. This indicates that the ritual of the interaction has the highest priority and that the meaning of the words spoken is less important than maintaining the ritual pattern. This view is supported by researchers in the field of study often referred to as phraseolology, where the formulaic nature of words and phrases is explored. Pawley & Hodgetts-Syder (1983) suggest that not only words but sentence stems are often memorized as one lexical item. A sentence stem is a clause length unit whose grammatical form and content is more or less
determined. While some elements can be changed those crucial to the stem remains fixed (ibid:190). Knowledge of these stems is crucial to be able to speak a language that is perceived as idiomatic. They argue that while many options may be open, which are all syntactically and grammatically acceptable, native speakers tend to favour only certain forms. As greeting phrases are considered highly formulaic (as discussed later in section 2.3) it is important to determine in what way they are allowed to vary and what could be considered their stem.
Wray (2002) prefers the term ‗formulaic sequence‘ as she points out that the current terminology is very inconsistent in its usage. She defines a formulaic sequence as: ―a sequence, continuous or discontinuous, of words or other elements, which is, or appears to be, prefabricated: that is, stored and retrieved whole from memory at the time of use, rather than being subject to generation or analysis by the language grammar.‖
8 (Wray 2002:9). Wray argues that, in contrast to Chomsky‘s idea that an infinite number of utterances can be generated by the speaker as long as they adhere to the rules of syntax and they form an utterance that has a recognizable meaning, only a handful of these possible combinations are used and considered idiomatic. She feels that we tend to prefer ‗prefabricated phrases‘, phrases we have encountered before, and therefore we say ―Hi, how are you doing?” rather than “What a pleasant event it is to see you. Tell me, how is your life progressing at the moment?”, even though both are possible (Wray 2002:12). Levin (Forthcoming:5) argues that commonly known formulaic expressions assists in creating quick production and comprehension, making them a very useful tool for communication.
2.1.3 The social aspect
The formulaic nature of language is one aspect that imposes limitations and rules for interaction, as discussed in section 2.1.1; another significant factor is the social and cultural setting. Language is sometimes used only for purposes of creating unity or solidarity, what is referred to as ―phatic communion‖ by Malinowski (1923:303). The aim in this case in not to inform or perform a more typical speech act such as a request or command, it is about being social and caring for relationships. In these cases there is often little meaning in the words spoken, the act of saying them is what is important.
Brown and Yule (1983:1-3) call this the interactive view of discourse as opposed to the transactional view, focused on communicating information.
Goffman (1967:306) likens speakers to actors giving a performance. He introduces the word ―face‖ which can be connected to the different masks an actor wears. To Goffman a speaker creates a face to project an image of who they want to be and how they want others to see them (a romantic, good employee etc.). This face is often constructed based on the person‘s idea of what society values (how a romantic or a good employee is expected to act. Since a person‘s feelings often become strongly connected to their face the evaluations that others make of their face have consequences.
An evaluation of face that is higher than the individual‘s own is likely to make the person happy and raise their self-confidence, thus affirming the person‘s position in society. A negative evaluation can lead to insecurity, sadness and may lead to a conflict.
An individual‘s behaviour needs to be in line with the face selected, because that face needs to be maintained (ibid:308). A priest may not behave in a way that raises questions about his faith for example. This responsibility according to Goffman binds
9 the actors in a contract to protect their face in every conversation. Since negative actions against others can be damaging to the performer it therefore also follows that the face of the person you interact with needs to be protected as well. Failing to protect or maintain or protect your own and your interlocutor‘s face can lead to conflicts or confusions since it means that the rules agreed on have changed.
Goffman therefore summarizes an individual as: ―a kind of player in a ritual game who copes honourably or dishonourably, diplomatically or undiplomatically, with the judgemental contingencies of the situation‖ (Goffman 1967: 317). Greetings are considered by Goffman (ibid: 318) to have three purposes:
1. To show that a relationship is still what it was, that it has not changed since the last time the interlocutors met.
2. To establish the roles the interlocutors will have in the conversation about to take place.
3. To state or apologize for circumstances affecting the relationship, such as apologizing for making someone wait or explain why it has been so long since they last spoke.
Holmes (1995: 338-343) discusses the differences in men and women‘s speech. She concludes that women in general tend to take a more cooperative approach to
interaction, using frequent and quick backchannels and statements of support or agreement, while men are usually more combative, challenging and insulting speakers etc. This also covers face-threatening acts (acts that may harm someone‘s face); women tend to a larger degree to soften statements that may offend and make them more indirect. On the other hand it is more common for men to baldly disagree. Holmes also points out that the differences between men and women‘s speech tend to decrease in mixed conversations indicating that they accommodate their speech.
The variation in women and men's speech is an issue that will be considered again in more detail in the next section that deals with a much debated aspect of discourse known as politeness.
10 2.2 Politeness
Politeness is a phenomenon that has received a lot of attention in linguistics during the last few decades. The connection between human behaviour, societal norms and pragmatic meaning may be some of the reasons for this interest. The study that has perhaps received most attention on the subject is Brown & Levinson (1987). In their book they propose a framework for categorising different types of politeness in discourse based on the function the utterances serve. The theory is based on two main assumptions. First that everyone, or in their words ―all competent members of society‖, have and are aware of the notion of face. The idea of face is based on Goffman's (1967) theory discussed in the previous section, in particular that all humans have a positive and negative face. Positive face is a person‘s self-image, something that they wish others can appreciate and like. Negative face refers to the boundaries that we wish others will respect and not intrude on. The other assumption is that we are all rational beings and strive to reach the most beneficial outcome of our interactions.
This model has been criticized by Watts et al. (2005) that argue that the general term politeness is a difficult concept to define and one that does not exist in all cultures.
Watts (2005:51) suggests that the term ―politic verbal behaviour‖ be used. Others like Janney & Arndt (2005:24) suggest more specific terms like social politeness and tact. In their definition social politeness represents the rules for what is seen as good behaviour in a culture or society while tact is on the individual level and is closely related to
―face‖ and the avoidance of conflicts in ―conversational negotiations‖. They base this on the idea that communication is often ―guided‖ by assumptions, cultural or social, about how the person we interact with will behave in different situations. Therefore if two interlocutors have different assumptions there is a possibility that conflicts can arise. In these cases Janney and Arndt (2005:31) note that it is usually not seen as a challenge but rather incorrect behaviour. Following this hypothesis it is then reasonable to assume that the choice of a form in a greeting ritual is influenced by social
assumptions with the intent to avoid conflict and show correct behaviour.
Another approach to politeness is Fraser's (1990:233) idea of the ―conversational contract‖. He suggests a more dynamic model where two people that enter into an interaction each have some notion of what to expect from the other, what rules they are obliged to follow and what rights they have. This contract can then be renegotiated during the interaction, if the context or relationship of the participants changes over time. Politeness to Fraser is all about operating within the limits of the contract;
11 politeness is simply normal behaviour. Therefore politeness is not noticed by others, only behaviour that steps outside the contract, rudeness, is noted. This suggests that the formulaic usage of language in politeness rituals may be influenced by a desire to not have your behaviour noticed as deviant from this ―conversational contract‖, which could be one possible explanation for the limited amount of forms used for greetings and predictable patterns found in the study in this paper.
Watts (2005:44) points out that the concept of politeness was originally more about signalling social standing. This may be one of the reasons why the term today has a slight negative connotation to speakers of English. Being used to regulate
memberships of social group‘s politeness is bound to change over time as social groups do. Also some tactics, such as warnings and threats, can be used for the same purposes as politeness strategies but are not considered polite. Due to this ambiguous meaning of politeness Watts, as stated earlier, prefers the term ―politic verbal behaviour‖, and considers it to be concerned mainly with ―maintaining social equilibrium in
relationships‖. Watts suggests five factors that determine ―politic verbal behaviour‖
(Watts 2005:51):
1) The type of social activity or setting 2) The speech events engaged in
3) ―To what degree participants have the same cultural expectations‖
4) To which degree the participants have the same assumptions about what information to share
5) The social distance or the power relationship
It is important to note here that a power relationship is asymmetrical. The person higher in the power hierarchy may use intimate language to those lower in the hierarchy but not the other way around. Meanwhile the social distance factor is symmetrical where close distance allows the use of intimate language and distance does not for both interlocutors. Which of these factors take precedence varies in different cultures and settings.
Watts (2005:52-58) considers the subject of universals in politeness to be difficult. While some linguistic features like forms of address are universals, whether
12 they are polite or not depends on the context they are used in. He illustrates this with examples from a radio talk show. It seems reasonable that since radio is a very public forum where the caller and host are not familiar with each other a more formal approach should be merited; however, since the host sets out to have a more cheerful and
interesting atmosphere he pokes fun at listeners who address him by his last name. In fact, throughout the show he constantly changes the way he addresses his listeners to throw them off guard (Watts 2005:62-68). In order to study politeness Watts believes it is necessary to approach it from two different angles. One is on the descriptive and comparative level and includes studying ―ritualised expressions‖, within which greetings can be counted, used in situations with different social distance and power relations to ―negotiate intimate stuff‖(personal information/exchanges). The second is more along the lines of Brown and Levinson's (1987) study and focuses on ―how the linguistic system can be used to symbolize the underlying changing nature of the exchange‖ (Watts 2005:58). In situations where neither social distance nor dominance are considered relevant factors, intimate language is the politically correct verbal behaviour. Like Janney and Arndt (2005), Watts (2005:62) also feels that it is bad judgement regarding which sociocultural factors are in play that leads to non-polite verbal behaviour, or face threatening acts.
Held (2005:131-151) notes that a lot of the work done in politeness is related to tactics of directness and indirectness, for example studies by Searle (1969), Lakoff (1973), Leech (1983) and Brown and Levinson (1987). Indirectness lowers the obligations on both conversational partners. It leaves the hearer with all the choices.
Indirectness also gives the interlocutors the chance to adjust or retract what they have said (Held 2005:141). Other work has focused on turn-taking, ritual order and other aspects of conversational analysis. Researchers like Edmonson (1981), Owen (1983) and Blum Kulka (1987) believe that ―conflict free relations and positive cooperation can be achieved with redundancy and explicitness‖ (Held 2005:133). Held considers a key feature of politeness to be that it is ―recipient design‖, in other words it is aimed at and often gives control to the hearer. She traces this to Goffman‘s (1967) ideas of
―supportive interchanges‖, which pay attention to and respects the interlocutors social personality, and ―remedial interchanges‖ that are face threatening utterances that might need to be weakened or re-worked to avoid conflicts (Held 2005:134-135). She also points out that politeness is not necessarily limited to utterances but can be expressed through structural combinations or complex sequences of actions (Held 2005:143).
13 Held also believes that the idiomatic aspect of language plays an important part in politeness, it is a part of our ―communicative competence‖ (Held 2005: 148). She claims that there is a relationship between ―politeness and linguistic routinisation and automatisation‖ (ibid). Held (2005:149) feels that we often use formulaic language and recurring behaviour to regulate social interactions and make the following remarks on rituals:
Biologically they are about regulation, control and integration used by for example pack animals (Callan 1970:80f).
They are a ―symbolic mastering of situations‖ (Hartmann 1973:139).
They are linguistic units that become formally fixed and thus separated from the original meaning of the individual words, instead they mainly have a pragmatic function.
They are used to regulate social encounters.
As greetings are usually considered highly formulaic in nature and are most definitely a required part of social interaction these ideas are likely to apply as well.
The previous research discussed in this section suggests that politeness, of which greetings are a part, is used to avoid discomfort and conflict in conversation and respect the interocutors‘ faces. Tactics like following the norm and being indirect are therefore seen as a common fall-back position. If politeness is influenced by social assumptions and used to regulate social interaction, it is plausible to assume that they will vary depending on the social context and depending on the speakers‘ relationship, which is something that will be discussed in the next section.
2.2.1 Politeness, class and social groupings
Seconding Watts‘ idea of the significance of history Ehlich (2005:73) also considers politeness to be something that is constructed historically and whose rules change with time. He remarks that historically politeness has a strong connection to high society, the court in particular, as a way of separating themselves from other social groups. Consider the English synonym courtesy or the German Höflichkeit (Hof = ‗court‘). Politeness was
14 developed alongside social change towards centralisation and follows the ideology of the ruling social class.
Werkhofer (2005:190) discusses the metaphor that politeness is like money based on observations on money by Simmel (1900). He draws the following parallels:
- It is a socially constituted medium (not just individual).
- It is a symbolic medium, derived from something else (values).
- It changes throughout history
- After a certain amount time the medium can take on power of its own. It is no longer only a means to an end but also something that motivates actions.
- In relation to the previous point the ability of one individual to master the medium will diminish the more power of its own the medium takes on.
Werkhofer also makes the following observations on social values in relation to society:
- The social values that politeness is based on come mainly from social order and social identity.
- These in turn are defined as the rights and duties that vary with the relationship and the interlocutors‘ positions in the relationship.
Structuring politeness is thus made more difficult since each person usually has more than one social identity and is a member of many different social groups. Therefore one of a person‘s identities may come into conflict with another, for example someone who is both a teacher and a student. So clearly there are many indications that politeness is connected to social standing and is therefore likely to change when social circumstances do. The study of greetings conducted in this paper provides empirical data on which greetings and greeting patterns are used by speakers in contemporary American English that can then be compared with past or future data. The next section will look more specifically at the social factors of age and gender and how they can influence choice of politeness strategies.
15 2.2.2 Politeness, age and gender
According to Holmes (2008:176) young people gradually start using more standard forms, which are usually more formal, usually peaking between the ages of 30 and 55.
She argues that the reason for this is that this is when people experience the highest pressure from society to conform. Since standard or formal forms are often considered polite this would suggest that the use of this politeness strategy is more likely with adults. Coulmas (2007: 61) agrees that the norm in language usually comes from adults.
He refers to this group as ―norm guarantors‖, commanding the highest societal strength, but also experiencing a higher degree of pressure to conform to societal norms. He argues that an adult speaking like an adolescent is more likely to be noticed, as that deviates from the norm, than if an adolescent speaks like an adult, which would be movement to the norm. Since adults are seen as a more powerful and dominant group in society, Coulmas (2007:64–65) argues that people are likely to be more polite to those that are older.
However, the more frequent use of one form by adolescents is not necessarily only indicative of the term being vernacular. Holmes (2008: 216) asserts that using the
―apparent time method‖ of comparing the speech of people of different ages can also provide clues as to whether language change is in progress. If this is the case than one could expect a steady decline or increase of a form the younger the speakers are. This is not to be confused with the pattern that indicates that one form is more common with the young and the old, which may indicate that this is a vernacular form.
Standard forms are also more commonly used by women while men use more vernacular forms (Holmes 2008:160). She suggests that possible explanations for this is that men as the dominant group do not need to use more politeness, that women are more status conscious, that women are expected to behave better or that vernacular forms indicate machismo. Coates (2004: 92-94) points out that women ask more
questions than men, but that many of these are from powerless speakers striving to keep the conversation going rather than take control and elicit an answer. In situations such as board meetings or seminars where questions are generally information requests men asked more questions (ibid), something that is very much relevant when analysing questions and responses in greetings.
In her study of 22 children between the age of two and five Gleason (1980:25) discovered that politeness is rarely spontaneous in children, but is very frequently prompted by parents with utterances like Say thank you or What do you say?. When
16 looking at the greetings hi and hello she found that there was one difference between boys and girls; the boys spontaneously said hi when the project assistant entered the room 44% of the time, but the girls only said hi 18% of the time. Gleason suggests that this difference can be explained by girls being shyer and that society puts more pressure on boys to provide greetings, shake hands or stand up if a woman enters the room.
Whether these traditions still hold true 30 years later can be discussed, but the empirical study in this paper also found that men more often initiated greetings and used more terms of address and additional questions like How are you? than women did.
2.2.3 Politeness and culture
One more aspect that deserves mentioning is that of the role that culture plays in social interaction. Blum-Kulka (2005) looks at the difference between the idea of politeness in America and in Israel, focusing her studies on speakers of Hebrew in Israel. She notes a stronger notion of ―frontstage‖, discourse that takes place in a public place, and
―backstage‖ which is the more private discourse occurring with friends of family. Blum- Kulka argues that while Israelis also have similar notions of politeness being a ―show of consideration‖ and a way to ―keep the peace‖ they also consider too much politeness to be flattering or manipulating. She notes that Israelis do not find American greetings and goodbyes very polite. Israelis feel they should be shorter and more sincere (Blum-Kulka 2005:265). She also believes that American parents when teaching their children
manners in speech focus more on turn taking while Israeli parents are more concerned with language correctness (2005:263). Blum-Kulka suspects that this is connected to the fact that Hebrew is a recently revived language and most speakers are either first or second generation native speakers. Mitigation in Hebrew discourse is more focused on gaining cooperation through stressing affect, using terms of address like dear or sweetheart for example. Other strategies are: Avoidance (not making face threatening acts), volubility (explaining your reasons), solidarity (being friendly) and being formal.
Ide et al. (2005:281-297) introduce the concept of ―wakimae‖ which is usually translated into ―discernment‖. This is the idea that some rules on politeness are highly conventionalized and therefore some responses are more or less culturally obligatory (discernment). In contrast to this the speaker usually has some form of choice in what strategy to choose (volition). Ide argues that in the Japanese culture discernment plays a bigger part while in western culture it is more focused on volition. This being said, both these phenomena usually play some role in any society. The view on politeness is also
17 different in that in English it is often closely connected to being friendly while in
Japanese it is sometimes more connected to formality and social distance.
Coulmas (2005:303) observes that the use of honorifics in Japanese is so encoded in the language that it cannot be compared to the western view of it as a strategic choice, but more of a norm. As an example of this he shows a transcript of a recording of conversations that the cabin crew had moments before the catastrophic crash into Mt Ogura in 1985 (ibid). Even though this was an emergency situation with high stress and a lot of shouting, the social relationships were maintained and the proper titles and politeness used. Coulmas argues that Japanese people tend to focus more on their relative position to the hearer rather than show concern for their individual face.
Which group you are a member of or represent (your company, family etc.) also plays an important part in the selection of the appropriate term of address to use. A secretary can refer to her boss by only his last name when talking about him to a customer
―Tanaka is not in today‖ but would never address him without using his appropriate title. This is because she is expected to be humble when talking about herself or the group she represents or is a member of.
Clearly these theories suggest that in any culture a lot of politeness is a combination of what is the cultural norm, expected or even obligatory, and the individual‘s choice of politeness strategy (volition). How much is up to volition and how much is culturally obligatory then depends on the culture, with volition having a stronger role in western culture. Since this study investigates patterns in choices of greetings in connection to social factors like context, gender, age and social class it says more about the cultural norms in society, though clearly volition can be used to explain deviation from these norms.
2.3 Greeting words and greeting questions
Greif & Gleason (1980) suggest that the phrase bye-bye is one of the earliest
communicative acts acquired by children. Their study suggests that greetings and other social routines are insisted on by adults. The children often only used them when prompted to do so by adults. When the usage is spontaneous it becomes clear that children do not really understand how and when to use them at first. In one of the examples the child was prompted to say goodbye to a lab assistant, but did so after that person had already left the room, unaware that it needed to be said before they left. This
18 early introduction of what Greif & Gleason refer to as social routines is a strong
indication of the importance we place on them. In the LSAC there are a number of examples of speakers discussing greetings, asking if someone was greeted, prompting someone to greet someone else or expressing their intent to go and greet someone. As for example in this conversation between three 16-year-old high-school girls:
(1) <1467> <laughing>She's mad at you <unclear></laughing>
<1469> Why? ... oh, my gawd, Tara
<1467> 'cause you didn't say hi when she said hi to you
<1469> Yes, I did, I did.
<1468> Uh hum. You never did. Not like <unclear>
<1469> She's just pissed because we hit her with water balloons.
<1468> Oh, but you didn't even really hit me. (LSAC)
Thus the knowledge of which greeting forms to use in different contexts that this study provides is important for people to successfully interact with others.
Previous studies indicate that there are different ideas about what a greeting is and is not. Duranti (1997:67) suggests six criteria to universally determine what is and is not a greeting.
1. Near boundary occurrence: An attention getting device, establishes a shared field of interaction.
2. Establishment of a shared perceptual field: Used as a recognition of each other's presence.
3. Adjacency pair format: The utterance made by one person, person A, creates an expectation of how the other person, person B, will respond.
4. Relative predictability of form and content: Despite their formulaic nature Duranti feels that we cannot assume that greetings have no propositional value.
He therefore argues that they are not completely predicable.
5. Implicit establishment of a spatiotemporal unit of interaction: A greeting marks the beginning of an interaction, how long this interaction is depends.
People who work together probably only great each other at the beginning of the day while people who run into each other twice in a day in different
circumstances may greet each other again.
19 6. Identification of the interlocutor as a distinct being worth recognizing:
Greetings can be used to identify class, group membership etc.
Duranti stresses that to study the full pragmatic impact of greetings the semantics of the words need to be paired up with ethnographic information. In other words the phrase or word needs to match up with the ―sociohistorical circumstances‖ (Duranti 1997:88). He suggests that they are always adapted to context and sometimes even assist in creating new contexts. Duranti firmly believes that the questions and answers posed during a greeting can also provide information, this he offers as a response to the notion that many feel they have little or no propositional content. Often someone higher in the hierarchy may request information. He insists that the choices made by the speakers can provide clues to a person‘s identity and origin. This study accepts Duranti‘s hypothesis and therefore criteria 1,2,3 and 5 were used to identify greetings in LSAC (see section 3.3) while the truth of criteria 4 and 6 are the subject of what this study investigates (see section 4).
Grimshaw (1980:793) suggests the following functions for greetings:
1) To show solidarity or respect
2) To obtain and validate presence, recognition
3) To introduce displays (to call attention to something such as the unexpectedness of the greeted person being in this location)
4) To threateningly show recognition (indicate that their presence is not welcome) 5) To ―finger‖ (identify) someone to others.
6) To reduce uncertainty or threat.
Criteria 3-6 do require rather special circumstances; in this study the first two criteria are more likely the functions used. The function a greeting has can be difficult to assess without detailed knowledge of the speakers‘ relationship. However, analysing speaker‘s choices of greetings based on their interlocutor‘s social belonging may indicate if they wish to show solidarity or respect.
20 Greetings can in some situations grant interpersonal access (Youssouff et al.
1976:812). Once gained, this access can then allow someone to make requests etc.
Grimshaw feels that while greetings have little referential value, they are subject to universal and cultural rules and can vary depending on the social context. Based partly on Goffman (1967), Youssouff et al. (1976) believe that there are conditions under which greetings are obligatory. If the two interlocutors know each other a greeting must take place unless there are special circumstances. If such circumstances do not exist, not greeting the person is, according to Labov and Waletsky (1967), considered a
―reportable‖ failure. It can be interpreted as rudeness, non-recognition or as a breach in the relationship. In the study conducted in this paper such conflicts were not detected.
However, there were examples of such breaches being discussed in the corpus, see example (13) in section 3.3.
Youssouff at al. (1976) study a very specific and in some ways extreme case of greetings, those that take place with the desert-living people called the Tuareg. As with Duranti (1997)‘s study these studies are slightly anthropological in nature and deal with cultures quite different from ours, they may however help identify universal rules for greetings. These hypotheses are then tested during the analysis of the data in this paper.
Sacks (1975:256) refers to greetings as ―ahistorically relevant‖, meaning that they occur even if people have greeted each other before, and marks that they should
―properly‖ appear at the beginning of a conversation. The study conducted in this paper agrees that this is the case unless a conversation is already taking place and an
additional speaker is approaching; in which case the new speaker is greeted, possibly provided relevant information about the conversation, and the conversation typically resumes. He believes that the most basic conversation is simply an exchange of greetings and calls this a ―minimal proper conversation‖. So an exchange that takes place can either be simply an exchange of greetings or an exchange of greetings and more. If greetings are left out there has to be a valid reason. He also points out that in cases where a greeting is not returned it is very often repeated until it is returned (Schlegoff 1968:117). Sacks also introduces the concept of what he calls ―greeting substitutes‖. These are questions, such as How are you?, that sometimes are used as greetings, but can also just be a part of a greeting, as in Hi, how are you?. A more detailed definition of how these questions are dealt with in this investigation can be found in section 2.3.1 below.
21 Sacks (1975:256) is quick to point out that questions such as how are you? are often not answered truthfully, nor should they be. His hypothesis is that interlocutors monitor the situation and choose an appropriate answer. So their responses could be positive (I feel great!), neutral (I'm ok) or negative (Not so good). However, while the monitoring suggests that the correct answer is a negative one, the interlocutor often chooses a neutral response instead. This is, according to Sacks, because negative responses always lead to a ―diagnostic‖ process where the other participant asks why the respondent is not feeling well. This process can touch on matters that are private and information which we do not wish everyone to have access to. Hence, if there is not a close enough relationship the respondent may lie to avoid giving away this information.
Sacks drives home the point that since the question is not seen as intrusive and the burden of deciding how much to share is squarely on the respondent, people will on occasion have to lie. It should be noted that positive responses may also lead to a
―diagnostic process‖, but can just as easily be ignored or responded to with a comment, like That's good to hear. Part of the study conducted in this paper looks at the
connection between greetings and response (section 4.2) and agrees with Sacks that these questions (How are you?) can be ignored with no apparent consequences, though most of the time a response is provided.
2.3.1 Definitions of the words and terms used in this study
Since the focus of this paper is on greetings and responses, where the greeting is initiated using the words hi, hey or hello. As these words are central to the study it is therefore important to take a closer look at how they are traditionally defined. The following definitions are taken from the Oxford English Dictionary On-line:
Hi: ―1. An exclamation used to call attention.‖
―2. A word of greeting. colloq. (chiefly N. Amer.).‖
Hey: ―A call to attract attention; also, an exclamation expressing exultation, incitement, surprise, etc.;
sometimes used in the burden of a song with no definite meaning; sometimes as an interrogative‖
Hello: ―1. Used as a greeting. Also in extended use.‖
‖2. Used to attract attention.‖
‖3. Used to express surprise or to register an unexpected turn of events.‖
―4. Used to answer a telephone call. In early quots. also used in initiating a telephone call.‖
22
―5. colloq. (orig. U.S.). Used to imply (sometimes disbelievingly or sarcastically) that the person addressed is not paying attention, has not understood something, or has said something
nonsensical or foolish.‖
According to these definitions hello is the only word seen primarily as a greeting, while hi is only used as such colloquially mainly in the US and hey is seen as simply a
discourse particle and interjection. As section 3.2 will later show hello is far more frequently used in the BYU-BNC (Brigham Young Universities adaptation of the British National Corpus) than hi, so this may be a reflection of British English.
Merriam-Webster‘s dictionary On-line, which is more focused on American English, offers the following definitions:
Hi: ―—used especially as a greeting‖
―—used as an informal way of saying ‗hello‘‖
Hey: ―—used especially to call attention or to express interrogation, surprise, or exultation‖
‖—used to indicate that something is not important, that you are not upset about something, etc.‖
Hello: ―—used as a greeting‖
―: the act of saying the word hello to someone as a greeting‖
―—used when you are answering the telephone―
As can be seen the definitions are pretty much in agreement in terms of hey and hello;
the difference is that hi is here considered primarily a greeting, albeit one that is more informal than hello. Someone reading these well-respected definitions may therefore draw the following conclusion regarding the usage of these three words as greetings:
Hello is the standard/formal greeting, hi is a more informal greeting used primarily in the US and hey is not used as a greeting. These definitions, as chapter 4 will show, are not fully compatible with the way that they are actually used in spoken American English.
As the previous sections have shown there are a number of terms floating around when it comes to greetings and questions like How are you?. As has just been explained Sacks (1975) refers to these kinds of questions as ―greeting substitutes‖, since they can sometimes be used instead of for example a greeting word like hi. The problem is that they are not always a substitute, in this study they frequently occur as part of a greeting uttered, for example Hi, how are you?. In this study it was decided to use more neutral terms and refer to hi, hey and hello as greeting words (as they are made up of one
23 word) and questions like How are you? and How‟s it going? as greeting questions, since they take the form of a question. Whether they always have the function of being a question is something that will be discussed later in section 4.2. The term greeting is used more generally for the whole utterance and greeting sequence for all the
utterances that make up the ritualistic initial exchange between interlocutors. How these were broken down and categorised is explained in the next section that looks at the material and method used in this study.
3 Material and Method
This section of the paper describes the material used during the study, how said material was gathered and the methods involved in analysing the gathered material. At the end of this section the limitations of material and methods used and problems encountered in the process are also disclosed and discussed.
3.1 The Longman Spoken American English Corpus
The main source used for this study was LSAC. The corpus consists of 717 text files in total containing about 5.1 million words (Leech et al. 2009:100), which are all
transcriptions of recorded conversations. The material for the Longman corpus was gathered by the University of California in Santa Barbara and features 1000 speakers of different age groups, gender, ethnicity and from 30 different states (Pearson Longman 2012). Though the corpus covers 30 states a lot of the conversations took place in California, making it overrepresented, and may therefore not be considered
representative of Spoken American English in general. Biber et al. (2006:248) in fact argue that there really is no such thing as a general language, since there will always be variation between different registers and between the language different people use.
While cultural background, which to some extent depends on geographical location, is a factor in politeness, the speakers‘ backgrounds are not documented in detail in the corpus. Nor is it varied enough in the sample for this study to provide a useful insight into how this affects speakers‘ choices.The sample selected should however be able to represent norms present in American society, since as all informants are speakers of English located in the US.
24 In the sample taken from the corpus for this study the age of the speakers also tends to fall between 15 and 65 with the other age groups being underrepresented considering the proportions of the general population, where for example 20% of the population is above 60 (Stern (2005). According to Stern the speakers in LSAC are divided into five different age groups all accounting for 20% of the total number of speakers following mostly the general demographics of the US, they are: 18–24, 25–34, 35–44, 45–60 and 60+. Stern does however point out that some speakers younger than 18 are represented in the corpus and are seen to be a part of the 18–24 group, but clearly these speakers do not make up a big group since the above age spans were chosen. The reason why the last group, 60+, is so underrepresented in the sample in this study (only 21 of the 463 speakers for which age was documented) is therefore surprising. It may be that the contribution of this age group in terms of greetings is lower, which is possible but again would be surprising since the difference is so marked (20% in the corpus but 4.5% in the sample) and nothing indicates that they would not also use greetings to start a conversation. It is also possible that Stern‘s figures have been rounded up for this group as they are an approximation or that the speech contribution of this age group is smaller than other age groups. At the same time the youngest age group 1–25
represented 38% of the speakers in the sample in this study. Age is an important consideration in this study, but (due to the issue just mentioned) the results are to some extent limited for older speakers. Stern or other sources investigated could not provide information on between which years the conversations were recorded, though the information on the individual transcripts in the corpus indicate that most of them are from the early 90‘s.
According to Stern (2005) the contributors in LSAC are 50% Male and 50 % female. However, Levin (Forthcoming), in his study of phrases used to excuse oneself to go to the bathroom, puts the number of utterances made in LSAC at 29% by men, 42% by women and 29% by unknown speakers. This would suggest that either one of these figures are inaccurate, or that though the number of informants is equally distributed the contributions by women are more plentiful. In the greeting database created for this study 35% of the utterances studied were made by men, 43% by women and 22% by unknown speakers. The lower rate of unknown speakers is easily explained by the omission of greetings between two unknown speakers discussed later in section 3.3. Utterances by women are consequently 1% higher in the database used in this study and utterances by men are 6% higher. However, this variation is small enough to be a coincidence.
25 Each of the 717 text files in LSAC is a transcription of a recorded tape. The speakers participating in the corpora were ensured anonymity and therefore any last names, phone numbers and addresses mentioned on tape were not transcribed (Stern 2005). At the top of each text file information about the tape and the speakers are provided, this information often varies in terms of detail when it comes to both setting and speakers as demonstrated by the examples below.
Textfile 111401 Textfile 113801
<HEADER_BEGINS> <HEADER_BEGINS>
<TAPE_#> 1114 <TAPE_#> 1138
<TAPE_TITLE> Dinner with my aunt <TAPE_TITLE> Eya Lewis 1
<DATE_RECORDED> 17-Sep-94 <DATE_RECORDED>
<TIME_RECORDED> 7pm to 11pm <TIME_RECORDED> to
<CITY_RECORDED> Bay City <CITY_RECORDED> Duluth
<STATE_RECORDED> MI <STATE_RECORDED> GA
<SETTING/BUILDING> private home, house <SETTING/BUILDING>
<SETTING/ROOM> kitchen, living room <SETTING/ROOM>
<EVENT_TYPE> face-to-face conversation <EVENT_TYPE>
<SCRIPTED?> unscripted <SCRIPTED?> -
<EVENT> we're visiting <EVENT>
<EVENT_DURATION> 5 hrs <EVENT_DURATION>
<RECORDING_DURATION> 4 hrs <RECORDING_DURATION>
Textfile 113801 Textfile 118101
<SPEAKER_#_2974> 2974 <SPEAKER_#_1279> 1279
<FIRST_NAME_2974> #Uhuru <FIRST_NAME_1279> #Rebekah
<NICKNAMES_2974> #Eya <NICKNAMES_1279>
<AGE_2974> 23 <AGE_1279>
<SEX_2974> F <SEX_1279>
<NATIVE_LANGUAGE_2974> English <NATIVE_LANGUAGE_1279>
<NATIVE_DIALECT_2974> <NATIVE_DIALECT_1279>
<DIALECT_STATE_2974> NC <DIALECT_STATE_1279>
<CITY_CURRENT_2974> Lawrenceville <CITY_CURRENT_1279>
<STATE_CURRENT_2974> GA <STATE_CURRENT_1279>
<OCCUPATION_2974> Cashier <OCCUPATION_1279>
<YEARS_OF_SCHOOLING_2974> 16 <YEARS_OF_SCHOOLING_1279>
<ETHNICITY_2974> Black/African-American <ETHNICITY_1279>
<CORPUS_ELIGIBLE_2974> Yes <CORPUS_ELIGIBLE_1279>
<VOICE_ID_TIME_2974> <VOICE_ID_TIME_1279>
<VOICE_ID_QUOTE_2974> <VOICE_ID_QUOTE_1279>
In the transcripts the speakers are identified by their numbers, for example<2974>; if the speaker is unknown they are simply transcribed as a question mark, <?>. Some paralinguistic features produced by the speakers, such as laughter, are also included in the transcripts. Pauses are marked with three periods followed by an indication for how
26 long the pause is in seconds and speech that the transcriber was not able to
hear/understand is marked as <unclear>. Other features of speech, such as tone of voice or intonation, are not marked in the transcript. In (2) below a few of the above
mentioned transcription features can be seen.
(2) <1781> <unclear> ... (9) so you want to do like a ... three, like a little combination for each corner?
<1780> Okay
<1781> Okay, like three or four steps <unclear>
<1780> Each person has three or four steps for
<1781> Yeah
<1780> Okay so how are we going to tur=, turn 'em? Just say turn?
<1781> [nv_clears throat] (LSAC Textfile 118101)
Compared to other large corpora, such as COCA (Corpus Of Contemporary American English) and the BYU-BNC, the Longman Corpus has the advantage of being a corpus of unscripted conversations that in many or most cases also provides basic information about the speakers and the context, which is vital to this kind of study. That being said this information varies significantly in the different tapes and as all speakers are informed about the conversation being recorded the conversations may be less relaxed and more polite than they normally would be.
3.2 The pilot study
The importance and nature of greetings have been discussed earlier on in this paper while looking at relevant previous research. Greetings initiate conversations and serve to establish or re-establish the social contact between two interlocutors (Duranti 1997:67). For any person interacting socially in society it is therefore important to be aware of how the ritual works and what language choices are appropriate for different social groups and in different contexts. A pilot study was therefore initially undertaken to identify common greetings and to ensure that the sample would be large enough to provide relevant results. The table below shows the number of occurrences of a few common greeting words and phrases in LSAC. Of course it is important to point out that in many of these cases the words are not used as greetings, but for some other function, such as a discourse particle.
