Stephen Few argues that Edward Tufte’s definition on the term “chartjunk” is too loose. Instead of using it to label all additional decorations and graphics that don’t contribute directly to the dataset being present, people should take the intentions of the designers and the intended messages of the charts into consideration. Few also argues that the study “Useful Junk?”, which were used to deny Edward Tufte’s opinion, is not valid because of its shortcomings in methodology. I totally agree with the author that different visualizations have their own purposes and target audiences, which may be independent to the datasets they are presenting. We should take those elements into consideration when we define what are “chartjunk” and what are not. The danger is if the viewers can detect those intentions as something subjective and something that are created by the designers to express their own opinions.