Introduction
Social Context= Power
Hypothesis= I predict that celebrity figures use power in language to portray a higher level of power compared to an average person, and attempt to build relationships between them and their audience.
This corresponds with the theory of Dan Clayton- Positioning and Stance.
In order to conduct this investigation various tweets were selected which were published by a celebrity and the average individual.
Methadology
I selected ten tweets using stratified sampling, five tweets from a celebrity and five tweets from an ordinary person.This involved taking every third tweet from the twitter accounts which avoided cherry picking and gave a wide variety of data. Furthermore my data had no obvious anomolies that would affect my results, and I did not know these people on a personal bias making the data reliable and unbias.
Benefits and limitations included the fact that these ten tweets were no where near representitive of the target population of twitter users and the tweets selected did not come from a very large pool. It may have been more reliable and respresentitive to take every 20th tweet from the users twitter. This also has further disadvantages as it is more time consuming to collect the data. Finally, I feel that ten tweets was no where near enuogh data to analyse accurately. As each tweet only has a maximum of 140 characters this made it difficult to analyse and annotate.
Analysis
So far in this mini investigation of language on twitter I have found that celebrity figures such as Katie Hopkins use more persnoal pronouns such as I, you, she, he etc. to adress their mass audience as indivuals. The use of synthetic personalisation 'you' helps Katie Hopkins to address her audience as individuals as it feels like they are the only ones she is speaking to. Furthermore, she is also found replying to her fans a lot which shows she does talk and adress her audience directly which is a positive but also demonstrates that maybe celebrities try not to be this authoritarian figure we impose them to be. Instead, they're more role models and celebrities like Katie Hopkins like to adress their mass audience individually by tweeting them. Katie Hopkins is a very opinionated individual which is good in her speech as she uses this to her advantage by influencing her opinion on others in her tweets. Lastly, Katie Hopkins also uses a number of empty adjectives by adressing her audience when replying to their tweets as 'lovely'. However, this language technique is more to do with the social context language and gender rather than language and power. On the other hand non celebrities tend to acquire more multimodality in their language to express emotion. For example, using sad faces to express the fact that they are poorly. Whereas Katie Hopkins tends to use connotations of emotion such as 'xxx' to represent kisses. However this again could in fact be an aspect of the social context language and gender.
Conclusion
I feel my hypothesis has been proved in one aspect and this is the fact that Katie Hopkins tends to use personal pronouns to address her mass audience as individuals and is talking to these people. On the other hand the non celebrity uses language in his tweets in general. He does not use any personal pronouns which shows he does not have anyone to direct his tweets to. Therefore this supports the theory of Dan Clayton and the fact that people in higher power attempt to build relationships with the writer. Whereas the non celebrity tends to use general language with no personal pronouns to show who in fact he is talking to. Due to the limit of characters istead of explain how he feels multimodality is used which implies he must not have much power. The male who is non celebrity uses multimodality in his tweets to express feeling however, the female who is a celebrity uses connotations of emotion such as 'xxx' to portray how she is feeling and furthermore she uses empty adjectives such as 'lovely'. This therefore makes me wonder whether I had chosen the wrong social context to study and my hypothesis should have been linked to language and gender, or whether I should have spent longer looking for my data and analysed more of it to come up with a better analysis filled with more language techniques and terminology.
