Introduction
The observer’s paradox is also known as the Hawthorne effect in social sciences. In sociolinguistics, it was first coined and observed by William Labov when he studied variation and style in speech. He designed the famous sociolinguistic interview meant to get information about language.
In this interview, he emphasized on four structural parts that the interviewee was requested to do. The first was to read a list of minimal pairs; which was followed by reading a list of words in isolation. Thirdly, the interviewees were required to read aloud a short narrative; and finally engage in a conversation about their lives. This paper aims to analyse issues that arise from observer’s paradox and attempt to present ways that sociolinguists can minimize their impact.
Observer’s paradox issues
Observer’s paradox as a problem lies in the process of investigating language style and usage among individuals through speech collection. In the process, a sociolinguist sets out to account in the best possible way about the variations occurring between what people say against what they believe they say. Interviewees perceive that their intelligence is being sort, and hence they alter their speech to suit these notions.
As a result, the sociolinguist as a field expert faces the problem of non-reality. He cannot discern whether what the interviewee is real or not hence the idea of unknowable. A double bind situation occurs since what the sociolinguist wants to know can only be known through observation which is subject to change. The interviewee speech performance is influenced by the presence of the interviewer hence observer’s paradox (Crowley, 2007).
The observer’s paradox issue has repeatedly been the subject of scholarly interest in ethnographic linguists. This has especially been in the study of language sensitivity in the social situation context. The sociolinguist who is also an interlocutor forms a critical component in the process since he is responsible for the interactional role.
A sociolinguist can only collect targeted speech on a daily basis in order to obtain reasonable amounts for analysis. This is meant to overcome observer’s paradox. The interviewer can only engage fully in the conversation as a participator at the lowest level. This means that his social relationship in the process allows for the most minimal effect to the linguistic data collected (Meyer, 2005).
There needs to be an effective strategy laid down in the form of roles and procedures that outline what the interaction roles are for the interviewer. As a fact, the team interviewer contradicts this role hence the name interlocutor. His relationship with the social group under investigation needs to be present before any interpretation of any linguistic behaviour is explained to seek for certain senses.
The observer’s paradox also throws into light the question of social class between the interlocutor and the informant. In cases where the two are of unequal status, the informant takes the higher side and thus this influences the information collected. This results to information that reflects self-importance or even looks down effect unto the interlocutor hence any attempt to analyse the data results to wrong conclusions (Stockwell and Trask, 2007).
In cases where the interlocutor is of higher social status, the effect to the informant would be either that of aspiration to illustrate same social status to that of the interlocutor’s or one that is too affected by inferiority effect. In these two scenarios, the interlocutor does not have the best information to analyse language in the social group since already it has been affected by his presence.
Another issue that arises in the observer’s effect is the problem of bilingual social groups. In such social groups, it is very difficult to maintain an interactional conversation between two people due to the process of code switching. Informants keep on switching to the other codes available hence this becomes a challenge to the smooth flow of the desired speech language.
At times, both the interlocutor and informant are of the same social group and ethnicity. In this case, code switching is influenced by the need to fulfil the cultural demands of politeness hence promoting code switching. Again this problem is aggravated by the problem of status equality between the two persons (Wei, 1994).
The issue of age is also another observer’s paradox area of concern in the field interview. The informants have their own views about age. If an interlocutor is of a lower age than the informant, chances of false linguistic data collected are high. This is even compounded by issues in gender variations.
The informants and the interlocutor may influence the responses in the conversations to suit a certain gender hence observer’s paradox effect. The informants may think that the interlocutor is interested in studying their knowledge hence they may give misleading data that result to flawed analysis of the language in the social context (Crowley, 2007).
Education is also an issue that is raised in the process of a sociolinguistic interview. Speakers who have lower levels of education differ significantly when compared with those with high levels of education. The education level of the informant and that of the interlocutor may at times be very different hence promoting a situation where the informant’s responses are impacted greatly.
The interlocutor at the same time may despise an informant who has very low levels of education. This may create an unnatural relationship between the two. This leads to production of unnatural speech that does not qualify to the intentions of the process of sociolinguistic study of language variations in social groups.
The observer’s paradox comes into play when the cultures of the informant and the interlocutor are at extreme ends. In this case, the informant may find it difficult to interview a culture that is superior just as it would be to maintain professional equality when dealing with one that is lower than his (Aarts, and McMahon, 2006). The other regards ethnicity in the sense that different ethnic languages may pose problems to an interlocutor interested in collecting linguistic data. This problem is seen in the context of social relationship or rapport that the interlocutor must create in order to set out a conversational setting for effective collection of data (Crowley, 2007).
Sociolinguists have attempted to minimize the effect of observer’s paradox trough different techniques and measures. Arising issues from such techniques include ethics in the process among others. Ethics means refers to accepted codes and standards of procedures in behaviour in a given set up or society.
Informants who give misleading information leading to incorrect recordings breach the ethics of truth and honesty at all times. The data analysed forth from such recordings becomes flawed due to such issues in the process on the part of the informant or interlocutor doing it knowingly (Kastovsky and Arthur, 2000).
Any attempts to collect data using hidden recorders breaches the questions of ethics since a sociolinguist cannot get any data secretly in his research. The speaker must know at all times that he or she is being recorded otherwise it becomes illegal in any part of the world. The fact that the informant is a friend or family does not give the right to record them secretly.
This is because it only leads to breach of trust resulting to damaged relationships. The attempt to decrease the effect of observer’s paradox emanates from the desire to present quality work rather than quantity. This means removing the human informants in the process of research, of which it is not possible (Stockwell and Trask, 2007).
Another issue an interview related to observer’s paradox concerns the questions and how to ask them. The process of interviewing is guided by questions that are structured to get the most natural display of linguistic data. Interviewers must learn the questions that must change the interview to the friendliest of all. The interviewer must thus keep in mind that besides being a researcher and a fieldworker he is also a speaker and a hearer in the conversation.
The interviewer must thus use strategies like volunteering personal experiences, reacting and responding to new issues as well as following conversations to where the informant directs them. The questions must be developed according putting in mind that they must elicit data and create a conversation setting (Aarts, and McMahon, 2006).
Questions that elicit emotions have been found to be favourable in some situations as opposed to others. Therefore, the interviewer must use his social skills to allow flexibility in order to determine a successful interview process. In terms of setting the roles, the interviewer should present an interview that encourages the interviewee to offer information through closer relationship rather than a distant one.
In this role, the interviewer sets up a setting which allows him to behave as a learner rather than an expert hence every question leads to another which is related. The interviewee must get interested in the informant in order to produce a relationship that is supportive and building.
Information disclosure in the field work is concerned with letting the informant understand the purpose of the research. This is usually not possible since quite a number of informants do not understand issues in linguists that are under study. In an interview, this is an important aspect to and it should be kept minimum.
When informants are told the purpose of the study, there is a possibility of a speaker’s influence is possible. This may lead to performance bias that can be conscious or unconscious hence the problem of informed consent (Kastovsky and Arthur, 2000).
The practicalities of the time range that an interview can last are also other issues raised by the observer’s paradox. Labov in Crowley (2007) wrote that it should last between one to two hours. Determination of how long it should take must be dependent on what forms the subject of interest in the interview. This is because different linguistic data demands varying amounts of time for considerable data recording.
Also, the issue of breaking familiarity with the interviewer in order to present a situation that is familiar determines the time and extent of an interview. In practice, interviewers use a repetitive strategy that involves the same informant for a period of a long time.
Conducting interviews in a repetitive manner, which spans for a long period, has often in practice been a measure to minimize the impact of observer’s paradox upon the process. Repetitive interviews create an environment that promotes conversation rather than an interview’s formal setting.
Accurate transcription of data recorded from the field work is another issue of concern. The difficulties arising from transcription are based on the fact that no account can include everything recorded. The transcription may lead to the inclusion of any amount of contextual data hence the potential infinite extent of the problem.
This problem is also aggravated by time and cost. In addition, the interaction of human and machine contributes to the aggravation of the problem. Further arising from this issue is the problem of the subjective interpretation to recorded data that only leads to the truth which is intended by the purpose of inquiry that plays the judge and the audience (Murphy, 2010).
The recorded data may be influenced greatly by background noise and interruptions that negatively impact on the quality and clarity of such data. This presents problems in transcribing phases as tedious time is wasted while the quality of the information gotten is susceptible to alterations and changes not meant to be present. Therefore, there must be efforts to change these issues in order to get the best process when recording information. The issue of transcription extends to the methods of collecting data (Thierberger, 2012).
The instruments used to record data have sometimes been observed to contribute to the observer’s paradox effect. This is when they attract attention that changes the informant’s natural language elicitation. However, at other instances, materials used to record data present the problem of applicability to modern data analysis.
These instruments may lead to time wastage due to their incompatibility hence also affecting the quality by falling below efficiency standards. On the same issue, the materials that are used to record may not be comprised in terms of quality and durability. This may necessitate the need for extra and constant care. If this lacks, the effective process of the interview may fail, and this leads to waste (Penke and Rosenbach, 2007).
Minimizing the effects of Observer’s Paradox
Sociolinguists have been concerned with reducing the effects of observer’s paradox in efforts to collect natural speech from an informant. As noted the observer’s paradox is brought about by the presence of the field worker, a recording device and the task itself and thence minimizing strategies must revolve directly or indirectly to these issues. These may include modifying the interview, the topics to be discussed in the activity, as well as the situations in the processes (Thierberger, 2012).
Stockwell and Trask, (2007) wrote that using protocols that engage the informants in emotionally involving responses, recoding events that are public and being recorded helps to minimize the potential of interference in the process of research. Labov in his quest for this solution observed that there was a need for the interlocutor to make process most friendly upon the informant. This was through participative observation where the informant becomes part of the conversation hence reducing the formality of the interview process.
Another approach requires the interviewer to enlist the number of informants to more than one (Thierberger, 2012). In the process, the participants engaged in more natural speech between themselves rather than with the interlocutor hence resulting recorded data becomes empirical. This strategy result in a casual speech where the relationship between the informants is not restricted by formality issues or other constrains of the stranger effect.
Another strategy for minimizing the observer’s paradox involves increasing the number of interviewer’s in order to improve the dynamics of recording sessions as well moderation settings in the process. The presence of two interviewers allows the generation of ideas and topics in a wide context hence make the process natural (Chambers, 2003).
A third strategy involves removing the interviewer in the process although this has issues in the recording process. Macaulay’s in Meyerhoff (2006) notes that research onto two pairs of adolescents who were left alone to engage in a conversation led to more data gathered quickly from a number of speakers as compared to the classic interview.
It was effective based on the fact that the interview took the natural, conversational approach. In this case, the interviewer was actively removed from the process although there are methodological issues to consider.
As already mentioned, overcoming the observer’s paradox may be achieved through repetitive friendly interviews. Such interviews maintain a rapport hence including familiarity in the process. The moment familiarity is included in an approach then the sociolinguist enjoys a process that records almost casual speech. This gives data which is factual and helpful in understanding the variations in language as espoused by different social groups.
Moderating questions to suit a certain setting, as well those that have been known to demand most natural display, is another way to minimize the observer’s paradox. Informants would respond to questions by giving insights on linguistics’ features of their language if they are correctly designed.
In addition, the amount of time that an interview should take is also an effective strategy since the interviewer can avoid time ranges that are tiresome or inappropriate. This helps to minimize the observer’s paradox on the informant hence resulting to the desired information. Sociolinguists should avoid disclosing the purpose of research study due to the problem of bias that comes with informed consent concept (Thierberger, 2012).
In terms of differences in ethnicity, it has been observed sociolinguists who use family and friends achieve corpora quickly and from a large audience. Sociolinguists may use interviewers whose ethnic orientation coincides with the informant for effective, natural speech data collection. This is workable based on the fact that an interview where the conversation approach is employed leads to success.
Age difference is also another issue in the observer’s paradox which sociolinguists must try to overcome. It can be solved by first laying down the procedures and regulations that should be followed in a research process. In this first approach, the researcher identifies his target in terms of ages and thus the deliberate choice of the interviewer must be put into concern in order to coincide with the average age of the target interviewee (Thierberger, 2012).
Education levels are another issue which sociolinguists must resolve to overcome consciously. Where education levels are varying the observer’ paradox can be overcome by applying the use of different languages in the process.
The application of such languages helps to involve informants in less-engaging language thus reducing the problem of the language barrier. The interviewer goes down to the level of the informant to take the position of a fellow companion. This is as opposed to an educated stranger who is aloof to the lowest social groups’ level of understanding and perspectives.
Another universally acclaimed way to minimize observer’s paradox effect is the use of participative observation. The sociolinguist in this case takes to anthropological dimensions by spending time in the speech community in order to gain the social closeness while at the same time investigating the language. This method creates familiarity although it has its problem in terms of interference and influence (Meyer, 2005).
The problem of transcription is solvable through the application of standard procedures in recording and transcription. These standards should be put in place in considerations of the study intended, as well the purpose. At the same time, this helps to decide on the materials that support long lasting storage of information.
This would solve problems of lost data due to wear and tear, and obsoleteness. The instruments used for such recordings should not be seen to contribute to the observer’s paradox effect. Thus, sociolinguists may utilize modern, efficient, as well as portable recorders in order to minimize this effect.
Conclusion
The paper focuses on the aspects that arise from the observer’s paradox. This is done from the perspective of the interviewer, the interviewee as well as the interview process. The three are involved in explaining issues like education, age, gender, ethnicity, and language. Others describe issues like data transcription and recording. After the issues, the paper checks the different ways each one of them can be solved by sociolinguists. This is in an effort to overcome its challenge hence reducing or mitigating observer’s paradox.
Reference List
Aarts, B and McMahon, MSA 2006, The handbook of English linguistics: Volume 36 of Blackwell handbooks in linguistics, John Wiley & Sons, Oxford.
Chambers, JK 2003, Sociolinguistic theory: linguistic variation and its social significance Language in society, Wiley-Blackwell, Oxford.
Crowley, T 2007, Field Linguistics: A Beginner’s Guide, Oxford University Press, Oxford.
Kastovsky, D and Arthur M 2000, The history of English in a social context: a contribution to historical sociolinguistics, Volume 129 of Trends in linguistics: Studies and monographs, Walter de Gruyter, New York.
Meyer, GP 2005, Synchronic English linguistics, Gunter Narr Verlag, Tubingen.
Meyerhoff, M 2006, Introducing sociolinguistics, Taylor & Francis, New York.
Murphy, B 2010, Corpus and sociolinguistics: investigating age and gender in female talk, Volume 38 of Studies in corpus linguistics, John Benjamins Publishing Company, Florence.
Penke, M and Rosenbach, S 2007, What counts as evidence in linguistics: the case of innateness, Volume 7 of Benjamins current topics, John Benjamins Publishing Company, Florence.
Stockwell, P and Trask, L 2007, Language and linguistics: the key concepts, Taylor & Francis, New York.
Thieberger, N 2012, Linguistic Fieldwork, Oxford University Press, Oxford.
Wei, L 1994, Three generations, two languages, one family: language choice and language shift in a Chinese community in Britain Volume 104 of Multilingual matters, Multilingual Matters, FranfurtLodge.