Limitations of Geo- Twitter

Fair analysis is not possible with automatic corrections. Many use twitter on their smartphones. Most smart phones provide automatic corrections. Some of these automatic corrections, do not identify the different varieties of English. Most smartphones or even computer automated corrector identify with American or British varieties of English. Hence, other varieties of English or the different sentence structures would get corrected leaving the data collector with a bunch of English language varieties similar to that of American or British. 

In the Geo-twitter analysis, the clipped words tend to have meanings other than ‘because’. Hence might have been wrongly roped in. For example, ‘cos’ could also refer to the mathematical word, cosine and ‘cz’ could refer to Czech Republic in other contexts.

Tweets are short, and provide little context for analysis. Just like fingerprints, every individual would have slight tint of difference in their usage of language. Even when a group of people may be from the same region and may have grown up in the same kind of environment there might be differences in the usage of language. Twitter only allows 140 words for each post. Twitter users try to minimize their use of the words as well. This gives the data collector minimum context for analysis.

Not all parts of the country might have access to the internet nevertheless, twitter. The map below shows that twitter usage is concentrated in city areas as compared to rural areas of Europe. Therefore it might be difficult to get data of the different varieties of English spoken in different regions of the continent. The data collection might only be restricted to the regions with WiFi and where people have twitter. These regions usually tend to be the urban areas where more educated people live. The data from such a source would only be a reflection of the language variety spoken by the educated urban population. However, if this data collection process is done in perhaps 10 years’ time when technological development and knowledge of twitter has reached the rural population it would be considered a fair collection and analysis of data.

Limitations of the Geographical location study

There are outlier languages which are spoken in the elevated land but do not use ejectives (Everett C, 2013) . It is also difficult to measure the languages according to current geographical location. People speaking languages which were initially from the lower altitude areas may have moved to  higher altitude areas during the different waves of migration and also due to globalization. A language might not only be spoken in one place but may be spoken in a few regions which have different geographical aspects. Thus reducing the credibility of this study.