simonshen
patreon
【國際關係深度評：經典復刻】 互聯網與「第二次印刷術革命」
🕑 Added 2025-05-26 07:01:02 +0000 UTC我們早前談過一本很有趣的書《公元3000年100大最有影響力人物》，在同一作者的前傳正史《公元2000年100大最有影響力人物》，有一位排名極前的人物不為史家重視，那就是令活字印刷術普及化的古騰貝格（Johannes Gutenberg）。其實，北宋的中國人畢昇才被視作活字印刷始祖，但古騰貝格把活字印刷改良及建立商業出版社，這是西方文明發展極重要一步。
這題目的專著倒有不少，例如倫敦政治經濟學院教授迪特馬爾（Jeremiah Dittmar），他曾比較印刷術在歐洲城市開始流行前後兩個世紀的經濟數據，發現雖然印刷術普及前後的經濟增長無明顯分別，但一個世紀後，擁有印刷及出版社的歐洲城市較沒有的增長快兩至八成。
經濟還是次要，活字印刷術最大貢獻是改變知識傳授方式及對自身文明想像，從而顛覆社會權力結構。
在印刷術普及前，歐洲的知識來源是教廷與傳教士，教廷不但成為社會的壟斷權力，也是知識界的唯一殿堂。當時普及教育是教會大忌，一來有思想審查，二來著作往往要以拉丁文書寫，其實也是保障既得利益者的手段。即使個別天才能著書立典，在欠缺印刷術的情況下，由作者孤本到有書籍流出市面往往耗費經年，學說自然難以流傳。
但這一切，都被活字印刷術改變。例如研究教會史的學者巴里．沃（Barry Waugh）指出，沒有古騰貝格，馬丁路德的宗教改革不可能成功，學說更不會透過大量宣傳品、小冊子及其翻譯的德文聖經進入尋常百姓家。從此教會不能再壟斷知識，專業階層與中產階級才成為農民、教廷以外的新興第三力量。
印刷術的另一衝擊，還有民族主義及本土主義的興起。這是《想像的共同體》的理據之一，只是常被忽略。這本書的作者 Benedict Anderson認為，民族本來就是建構出來的概念，而印刷術及大眾媒體的出現，令民眾對於民族成員的「想像」，變得更為實在。例如共同的歷史想像及生活經歷，可以透過書本、報章，傳遞到「共同體」內其他成員；地方語言及文字，也是透過印刷術及書本得以流傳，形成屬於特定社群的共同文化圈；公共知識份子著書立說，挑戰「大一統知識霸權」，才讓民眾得到「教徒」以外另一政治身份的可能性。
說了那麼多歷史，有什麼現實意義？答案自然清晰不過。今天的互聯網革命，實際上就是第二次印刷術革命。教廷不能壟斷知識後，取而代之的是專業學院，特別是大學，而學術語言及種種離地術語，就像教廷的拉丁文，「影響因子」就像《十戒》，畢業證書就像教籍。
但互聯網普及後，一般人都可接觸從前大學才能傳授的知識，很多學院內的「禁忌」思潮都在互聯網大行其道，而從前的精英階層卻面對愈來愈大的挑戰，互聯網世代有自身的精英定義，正隱隱然衝擊着整個社會結構。
幾乎可以肯定的是，大學的壟斷就像中世紀教廷對知識的壟斷那樣，不可能在一百年後繼續存在， 那麼這一百年內，整個「反壟斷」過程究竟是馬丁路德那樣轟轟烈烈的革命，還是慢慢的軟着陸？局中人未雨綢繆，絕對是必須的。猶如蒸汽機的出現，很可能改變人類社會沿用數百年的生產模式，這又是一場工業革命。大家在讀歷史的同時，其實也正活在歷史。
*改編自沈旭暉《信報財經新聞》文章
Comments
Tisana limone
De Bende van Jan de Lichte, a tv series which mentions a father and his daughter were demonised because they own a press in 17th century, and their freedom was restricted.
堅離地書院 College
in the long run, church lost dominance because of that the trend might not be the same for AI's future
lyk
Technology changes but people don’t. Student A who has invested heavily in crypto will ask an LLM to go through 10000+ academic papers to help him find convincing arguments to support liberalising crypto. Student B who thinks crypto will destroy human civilisation will do the opposite. If the first LLM does not give them what they want, they will get another one that does. If not, train one yourself that gives you that conclusion.
Unsung
AI has been revolutionising the way we obtain data, process data and produce data, just like the Internet and information revolution did in the past decades. While the SOTA LLM/AI still suffers from hallucinations, it still emables human beings to chunk/digest 1000+ articles/papers on an issue within half an hour. Processing this amount of knowledge/information was unimaginable until super long context LLM was made accessible in the past year. I think it is no kidding that, after 10 years, any college student can say "my conclusion is based on scrutising 10000+ academic papers".
lyk
But we must also mention that one of the first things the printing press delivered is witch-hunt, and that lasted a loooong time. “The Hammer of the Witches” is one of the bestselling books, while books about dull mathematics and physics are not. Professional witch-hunters proliferated, data amassed, along with protocol of trials and “documented confessions”. The number of conspiracy theories far outnumber the number of truths, and they easily spread as humans are far more susceptible to outrage and misinformation than rational analysis. 「共同體」 — today the internet enabled a global interconnected network of limbic systems (emotions) rather than prefrontal cortex (thinking), so to speak. Universities have their flaws, but who then would be the authority on the truth — would that be Elon’s TruthGPT (AI), or Trump’s Truth Social (personality cult)? Neither sounds promising. Unlike the medieval Church, the hallmark of scientific institutions is self-correction, you pursuit the truth but you don’t claim to know the truth, and you are allowed to be wrong. That still serves an important purpose. AI is not the answer, if Elon does not want to be seen as fallible, his AI will be fed data that make him look infallible.
