SMS or text messages are usually seen as short and efficient, but also flawed with spelling errors, numbers and other visual symbols such as smileys. Relatively little is known about what exactly users do when they text and how much variability may exist between users. Recent Canadian studies have also overlooked the relationships between the language used in these messages and the users’ attributes. Investigating this novel use of English (and possibly other languages) in a large corpus will provide insight into how innovative and productive users can be even -- perhaps especially -- when technology imposes severe constraints such as a very limited number of characters. The objective is to gather a database of 100,000 text messages!