Magister Artium - MA / Social networking sites are used on a daily basis, to communicate with friends we have known for quite some time as well as make new friends from all over the globe - a global phenomenon. According to Aparicio (2011) the use of social networking sites have given way to a new “social dynamic” where friendships are formed with individuals from diverse backgrounds and geographical locations. Larsen (2007) continues to state that social network sites make for the creation of new friendships and the maintenance of new relationships. Using social networking sites to develop relationships provides us with new social skills, but through constant use of these sites we lose valuable interpersonal skills learnt through the use of face to face interaction (Aparicio, 2011). This thesis investigated the notion of social networking sites, specifically focusing on interpersonal relationships and language use within the networking context. The social networking site in question is that of Twitter, as majority of existing studies in this area focuses on the more popular Facebook. The main objective was to determine whether social networking sites, specifically Twitter, influence the development and maintenance of interpersonal relationships and language use. Participants included a group of 11 males and 11 females (22 in total), ranging from the ages of 17 to 33 and from different geographical locations (e.g. United Kingdom, South Africa, Tokyo, and so on). These participants frequently keep in contact with each other. Four ways in which Twitter has been adapted to emulate face to face communication have been found, namely: (1) the use of Paralinguistic and Prosodic Features to imitate speech, (2) Ustream, although not prominent within the data collected, is used to make up for the lack of face to face communication.This, however, is one-way; only one user provides a video link while those communicating with him or her (as there can be more than one) would type messages, (3) as expected a variety of shortenings can be found within the data collected. Shortenings imitate speech among the younger generation, and (4) participants make use of an informal register, as the most common type of relationship found on Twitter is that of friendships. Both strong and weak ties exist in the collected data; with weak ties being the majority. It is possible for weak ties to become strong ties. All online relationships start off as weak and gradually, over time, become strong ties. This is done through establishing trust between participants and communicating on a regular basis. Paolillo (1999) found that online relationships manifests as both weak and strong. However, “online ties are not ‘branded’ as weak ties” and these ties differ in quality; “those who have regular contact have strong ties and those with less frequent contact have weaker ties weak”. Social support is evident in the collected data and possible in online, textbased communication. In is manifested in four types of support, namely: instrumental, emotional, informational, and appraisal. The most common type of support found in the collected data is that of informational support. With regards to support activation strategies, most tweet fall under the ‘ask’ and ‘cry’ types of strategies. Also, considering the amount of emoticons found in the data, little or no emoticons were found in the activation strategies. There are also more indirect support activation strategies as opposed to direct. This could possibly be due to the fact that majority of the ties are ‘in the weak stage’. Textese has not been adapted in anyway; the same elements used by texters and IMers are used by tweeters, such as initialisms, phonetic spellings and contractions. Although present in the Twitter data, elements of textese did not occur as frequently as that found in e.g. Bieswanger’s (2007) and Thurlow’s (2003) studies; however more elements of Twitter language was found. If anything, the characteristics of textese are well-suited for Twitter; as shortened forms of words would make it easier for users to maintain a character count below the imposed limitation and it promotes the idea of writing quick and concise messages instead of filling message space with irrelevant content. With regards to the difference in the way male and female participants use language in terms of the linguistic characteristics of textese and the language unique to Twitter and the use of paralinguistic and prosodic features, it can be said that females tend to use these characteristics more than males do.