譯/莊蕙嘉、核稿/樂慧生
美國人按讚假新聞 4年變3倍
During the 2016 presidential election, Russian operatives used Facebook, Instagram, Twitter and other social media platforms to spread disinformation to divide the American electorate. Since then, the social media companies have spent billions of dollars and hired tens of thousands of people to help clean up their act.
2016年總統大選期間，俄國特務利用臉書、Instagram、推特和其他社群媒體平台，散播假訊息以分化美國選民。自此之後，社群媒體公司已花費數十億美元，雇用數萬人以協助清除此類行為。
But have the platforms really become more sophisticated at handling misinformation?Not necessarily.
但是，這些平台在處理不實訊息上真的變得更為老練了嗎？不盡然。
People are engaging more on Facebook today with news outlets that routinely publish misinformation than they did before the 2016 election, according to new research from the German Marshall Fund Digital, the digital arm of the public policy think tank. The organization , which has a data partnership with the startup NewsGuard and social media analytics firm NewsWhip, published its findingspublished its findings earlier this month.
根據（美國）公共政策智庫「德國馬歇爾基金會」的數位部門「德國馬歇爾基金數位」的新研究，比起2016年大選之前，民眾今日在臉書上更常接觸慣常發布錯誤訊息的新聞網站。這個組織和新創企業NewsGuard及社群媒體分析公司NewsWhip是數據資料合作夥伴，本月稍早發表了前述結果。
In total, Facebook likes, comments and shares of articles from news outlets that regularly publish falsehoods and misleading content roughly tripled from the third quarter of 2016 to the third quarter of 2020, the group found.
該組織發現，整體而言，慣常發布錯誤與誤導性內容的新聞網站所發布的信息，在臉書上獲得的讚、留言和分享數，自2016年第三季至2020年第三季約增為三倍。
About two-thirds of those likes and comments were of articles published by 10 outlets, which the researchers categorized as "false content producers" or "manipulators." Those news outlets included Palmer Report and The Federalist, according to the research.
這些讚和留言約三分之二是針對來自10個新聞網站的報導，研究人員把這些媒體歸類為「不實內容製造者」或「操縱者」。根據前述研究，這些媒體包括帕爾默報告和聯邦主義者。
The group used ratings from NewsGuard, which ranks news sites based on how they uphold nine journalistic principles, to sort them into "false content producers," which repeatedly publish provably false content; and "manipulators," which regularly present unsubstantiated claims or that distort information to make an argument.
這個組織使用來自NewsGuard的評分，依據新聞網站遵循九項新聞守則的程度加以排名，將它們歸類為重複刊登可證實為誤之內容的「不實內容製造者」，以及例常提出缺乏根據的說法，或扭曲資訊以支持其論點的「操縱者」。
"We have these sites that masquerade as news outlets online. They're allowed to," said Karen Kornbluh, director of GMF Digital. "It's infecting our discourse, and it's affecting the long-term health of the democracy."
「這些網站假扮成網路新聞機構，它們被允許這麼做。」馬歇爾基金數位的執行長凱倫•康卜勒說。「這正在毒害我們的公共論述，以及民主政治的長期健康。」
Kornbluh said Facebook users engaged more with articles from all news outlets this year because the coronavirus pandemic forced people to quarantine indoors. But the growth rate of likes, shares and comments of content from manipulators and false content producers exceeded the interactions that people had with what the researchers called "legitimate journalistic outlets," such as Reuters, Associated Press and Bloomberg.
康卜勒說，臉書用戶今年更常接觸所有新聞機構的文章，因為新冠肺炎疫情迫使人們待在室內隔離。但是用戶給予操縱者和不實內容製造者所提供信息的讚、分享和留言的增加率，已超過用戶和研究人員所稱的「合法新聞機構」之互動程度，路透、美聯社和彭博資訊即屬後者。
說文解字看新聞【莊蕙嘉】
資訊愈發達，難以查證或以假亂真的「假新聞」也隨之增加。最常見也最白話的用字是fake news，本文則用了兩個不同的字來探討社群媒體假新聞亂象。
本文標題所出現的misinformation，字首mis意為「no」，泛稱非故意或不確定意圖的假新聞；這類新聞經查證為假消息後，會以hoax稱之。第一段的disinformation，字首dis意為「wrong」，特指刻意捏造與傳播的不實訊息，帶有deliberately to deceive的意圖。
因此，第一段談到四年前美國大選，俄國透過社群帳號刻意散播假新聞，用的字是disinformation。之後的文章談到傳播不實新聞的「現象」，則用misinformation。
