https://bobostory.wordpress.com List

Sunday, November 26, 2017

专访许煜:人工智能的超人类主义是二十一世纪的虚无主义


轉載自: http://www.thepaper.cn/newsDetail_forward_1718074
学者许煜任教于德国吕纳堡大学哲学研究所,并负责该校的 “参与的技术生态学”研究,他同时是法国西蒙栋国际研究中心的研究员,中国美术学院跨媒体艺术学院的客座研究员。
许煜是法国哲学家斯蒂格勒的弟子,和后者一起合作了约十年。在法国哲学家、人类学家拉图尔“重置现代性”的工作坊中,许煜被邀请为中国场的嘉宾之一,响应拉图尔提出的人类纪困境及解决方法。
多年来,许煜关注社会运动和对全球化的反抗,如今却很少谈社运了,“我们必须思考为何左翼总处于下方,或者我们需要的是对资本主义的新的理解,来发展出新的策略”。
有一种刻板印象认为,做社会运动的都很反抗技术。许煜正是要反对这一对立,提醒人们把更多注意力放在技术问题。在他看来,欧美左翼的问题在于对技术的漠视,“结果很多人用Facebook来组织反资本主义的社会运动,就像用微信来对抗阿里巴巴一样”。在cyborg逼近的时代,他提醒人们警惕超人类主义,因超人类主义本身是一种更刺激的消费主义,而在现代性危机面前,每个文化都要重新发明自己的宇宙技术论。
以下是澎湃新闻对许煜的专访。
澎湃新闻:您现在主要在做的是“人类纪”这一块?
许煜:我的新书《论中国的技术问题》(The Question Concerning Technology in China. An Essay in Cosmotechnics, 2017年初出版)就是在讲这个问题。人类纪指的是人类的活动已影响了地质化学(geochemical)活动,那是地球在11700年前开始的全新纪或全新世之后发展的一个新的纪元。这也是全球的技术化、科技的全球化所带来的一个巨变,这给了我们一种迫切性来思考整个现代化以及现代性的问题。如我在回应拉图尔时所指出的,我们必须分开现代化和现代性。后者是十七世纪在欧洲出现的一种认识论和方法论上的断裂,而现代化是这些新的知识体系在地域上的扩张。
这种现代化同时也是单边的全球化。然而这种单边的全球化已到了尽头,近年来在欧美涌现的极右运动以及新反动主义(neoreactionaries)是最好的标志。如果您细看他们的论述,那是反启蒙的,也就是说反普世主义以及普世价值,如自由、民主、平等等,所以他们的言论充满种族主义、法西斯主义的味道,他们认为黑人与白人是不同的,现在将他们一视同仁是政治正确,也因为这些普世价值的政治正确,欧美的发展一直都匍匐不前。问题是,西方如何保持它的优势,就好像硅谷的著名投资者、特朗普的支持者,同时也是新反动主义者的彼德·泰尔(Peter Thiel)在911之后问道, “有没有一种不彻底毁灭现代西方,同时仍能巩固其地位的方法?” 我认为这种自觉是黑格尔在《精神现象学》中所说的“苦恼意识”,也就是说意识到自身的矛盾,但又超越不了这种矛盾。
很多左翼会强调普世价值来响应,但这是不够的,事实上我们不能忽略右翼论述的吊诡之处,他们或多或少指出了当前这种全球化的问题。如果我们从相同出发, 最后我们发展出来的就是今天被视为失败的多元文化主义(multiculturalism),其实背后还是单边全球化,我们需要强调的是差异,以及思考这种差异如何能够扭转当前的局面,而我在做的就是通过解构技术在人类学意义上的普世性,来重申差异 。现代化造成了dis-orientation,它不只是失向,而且是失去了东方性,也就是说误以为东西方是同一根源,失去了历史性的思考。我认为我们应该比那些右翼和新反动主义者走得更远,从绝对的不同开发出相同,而先明白中国跟西方的不同,最终才能达到如拉图尔和其他关注人类纪的哲学家和人类学家所说的共存(co-existence),也就是说文化的多元化成为整个全球化里的核心,而不是表面。
表面的全球化相对的是传统的实体化(substantialisation),将传统实体化表现在两方面。首先在国家治理层面,比如说通过它来定义某种民族主义甚至法西斯主义。第二个是文化工业,例如建造出很多创意、旅游小镇,把传统实体化成旅游纪念品。这是全球化的症状,它所强调的不同都是表面上的装饰性的东西。那问题是我们怎么从不同发展到相同?首先必须要强调不同,其次必须将传统去实体化。传统是一直处于变化之中的,从先秦到宋明,传统的精神保留了下来,有些是有用的,可以支持我们解决当前的问题,但是答案不是已经在那里了,答案是要被发明的。
面对当下复杂的全球化进程,我在《论中国的技术问题》中提出技术的多样化,不只是在中国,其他文化也应该发展出其技术思想来抵抗西方科技的全球化。但是,我并不是说要否认西方的科技,这是不现实的,而且中国也一直缺少对西方技术的哲学性的思考。我的想法是如何扭转这种发展的方向。传统可不可以以另外一种新的方式回归,回归不只是一种装饰,我们可以将它变成一种批判理论。我们现在所说的批判理论,不是德国的就是法国的。所谓霸权,有时候是无意识自己加在自己身上的。
澎湃新闻:我们从过去寻找来的思想资源,跟现在的科技怎么结合?
许煜:这就是我为什么提出首先要系统性地分析中国技术思想的原因了。从先秦哲学开始到至今,我们怎么理解器和道之间关系的变化。有些哲学家也看到这个问题,譬如说李三虎先生也强调器和道的问题,而且提出重申传统,但我想更系统化地理解器道的关系问题以及从认识论和知识型的角度来思考这一系统化分析如何响应当前人类纪的问题。首先,我们的知识论在现代化过程中都是被边缘化的,那么问题是这种知识论可以怎样被重新搬进来。其次,我视知识型问题为sensibility(感知性)的问题,也就是说它是不同知识领域的基础和链接。福柯在谈到知识型(episteme)的时候,他覆盖了三种,就是文艺复兴时期、古典时期、现代时期。但我们不能把福柯照搬过来放在中国,他为我们提供的是方法论。福柯是一个历史学家,他是一个思考历史的人,历史总是独特的。而哲学总是需要去回答所处时代的问题。但不少研究福柯的人都忽略了,这总让我想起海德格尔在《论人道主义书简》(Brief über den Humanismus )所说的,现在的人不再思考而要忙着搞 “哲学”。
在我提出的框架里,知识型的问题必须被重新思考。知识型的转变总对应着一个时代性的转变,也就是说当有一个历史性的危机出现的时候,我们就会发展出新的知识型来对应。我们可以归纳一下三个可能的知识型的转变。周室衰退时,我们就有了儒家的 “复礼”,道家的“自然”等,这就是重新发展出知识型来回应那个时代,一直到汉代,特别是在董仲舒那里一种儒家与黄老结合的天人观也固定下来了。而在唐代末期当佛教成为主导的时候产生了另外一个危机,由宋代理学所代表的道德宇宙论也出现了来作为响应,在公元十一世纪,我们见到宇宙发生论(cosmogony)与道德观以更严谨的形式结合。而第三个阶段是在两次鸦片战争之后的整个现代化运动中,比如说魏源、维新派,也都是对时代的反应,他们想要生产出“中学为体,西学为用”的知识型,但最终在西方科技的冲击下沦为了完全的西方化。理解这个过程是很重要的,它容许我们避免重复错误同时理解我们文化内里的创造力。
澎湃新闻:那您认为这次的西化和现代化是跟历史上一样的知识型转变,还是一次史无前例的断裂?
许煜:两次鸦片战争之后出现了那时处理不了的问题,因为对西方的技术一无所知,他们认为西方的技术就是中国哲学中的“器”。那为什么上两次(先秦哲学和宋明理学)会成功?宋代理学完全不涉及技术这个问题,尽管张载的哲学影响了后来宋应星《天工开物》技术和自然不分开的观点,而相对的,在《天工开物》之后一百年在法国出现的《百科全书》则是提倡自然和技术分开的(这也是鲁索对百科全书主义者的批判)。宋明理学其实是自然哲学,它发展出一套新的道德宇宙论。为什么在公元11世纪,宇宙发生论突然成了主要的学术理论呢?突然之间人们开始关心宇宙是怎么产生的了,这很有趣。因为他们是回到以自然为核心道德宇宙论,与技术无关。准确来说,构成这个知识型的不是技术问题。两次鸦片战争之后,我们被逼着去实现现代化,被逼着我们去发展船坚炮利,那个时候还不知道什么叫技术,都统称为“科学”,康有为统称为“物质学”。
回到您之前提到的问题,也就是说传统和现代科技怎样结合的问题,斯蒂格勒为我们提供了很重要的参考。他是一名真正的思想家,我这样说不只是因为他曾是我老师,而是他一直都在坚持,一直都在尝试,他的创造力是很让人佩服的,而且他很慷慨。我指出如果现代性所标志的是一种技术的无意识(technological unconsciousness),那么现代性终结的标志的便是一种技术意识(technological consciousness)。这种意识首先体现在利奥塔1979年出版的《后现代条件》,但系统性将这梳理出来的是斯蒂格勒,他的硕士论文其实是跟利奥塔写的。
首先,我们可以从神话以及宇宙论的解度来切入。如果斯蒂格勒说技术的来源是普罗米修斯,那是因为他是在欧洲的学术源流里这样论述的,但如果我们跟着说中国的技术起源是普罗米修斯那就好笑了。哲学的思考无法脱离神话,斯蒂格勒之所以能发展出哲学理论,是因为他将神话变成了他思考当中不可缺失的一部分。在1795年的德国流传下来一篇文章,题为《德国观念论最古老的体系纲领》,作者不详,据说是谢林、黑格尔和荷尔德林,或是这三名同学的其中一员,这是作者(们)有见于形而上学终结提出的哲学的发展方法,文章结尾说神话要有哲学性,而哲学要成为神话 (die Mythologie muß philosophisch werden…und die Philosophie muß mythologisch werden)。1795年,自然科学在欧洲开始变成一个系统,这时他们提出了再神话化,这是他们对所处时代的回应,也是对古希腊文化的延续,他们也回到了悲剧(das Tragische)。所以,神话跟哲学思考和文化的发展是分不开的,如果要出中国的技术思想来回应当前的问题,我们必须认真地处理我之前提到的三种知识型的转变以及道德宇宙论如何与技术发展重新结合,康有为他们很勉强地将中国的道德宇宙论以及西方的物理学等同,譬如说将以太和“仁”等同,这是非常错误的,我们一定要避免这种寻找等同(equivalence)的做法。
那么我们必须以另一种方式来提出问题,也就是说不先求等同进而吸纳,而是我们先从绝对的不同,比如说宇宙论和器道思想出发来找出路,重新将感知性纳入进来?其实,“道”也就是sensibility的学问。与康德不同,新儒家的牟宗三认为在中国哲学当中,人是有智的直觉的,他的感知性是超越现象的。牟宗三所说的“无限心”就是感知性的问题,只有无限心的感知,才能进入“物自身”。牟宗三从而发展出所谓 “良知自我坎陷”来展示无限心与认识心的转换的可能,从而证明中国也可以发展出技术,但这仍然站在观念主义者的立场上。当前中国已经发展出技术来了。所以我想要处理的不是这个问题, 而是在“中国也可以”的基础上思考技术本身的问题,即摆脱观念主义的束缚,将技术问题纳入到我们的体系中来。
一个是从它的内里来改变它,因为当下技术的发展是一个巨型技术系统,特别是大数据、智能城市、智能手机形成的大型技术系统,我们要从认识论的角度思考怎样从内部改变它,我和斯蒂格勒合作了很多年,一直都在研究就是这些问题。第二个是从外部,这是我所说新的宇宙技术(cosmotechnics)的框架,宇宙技术的翻译也是一个问题,拉图尔笑我说这两个字kosmos 和 technē都是希腊文,但如果我说“天人技术”又只局限在中国了,这是全球化下语言的局限,我想强调的是不止在中国,其他非欧洲文化也应该思考宇宙技术的问题。宇宙技术的基本定义是通过技术活动来统一宇宙秩序和道德秩序。我提出宇宙技术是要克服二十世纪以来西方技术概念的局限,如果我们可以这样归纳,在二十世纪西方技术哲学里有两个主要的认识,这也是海德格尔一九五三年的《论技术问题》(Die Frage nach der Technik)所留下来的,首先是古希腊人所说的τέχνη,它的本质是有诗意地带到跟前,海德格尔用了希腊文的poiesis和德文的 Hervorbringen来描述这个亚里士多德的四因说里的动力因(causa efficiens)。其次就是在欧洲现代时期出现的科技(moderne Technik),它的本质已不再是古希腊意义上那样的,而是所谓的座架(Gestell),也就是说所有的存在都被视为可支配的资源。但这两个概念并没有办法解释古代中国的、印度的、南美部落的技术。所以我发展出cosmotechnics这个概念来包容不同的技术,同时将宇宙论和技术重新结合,也就是说技术的发展总与它所处的文化的宇宙论紧密相关的,我们甚至可以说宇宙论的存在模式就是宇宙技术。在西方,宇宙论已经变成天文学或者天体物理学了,本来是背景(ground)的,现在完全地技术化、数学化了,人类纪标志的正是这种宇宙论和形而上学(海德格尔认为尼采是最后一位形而上学家)的终结,所以我认为要响应人类纪必须要回到多样的宇宙技术,这必然要重新启动传统以及其宇宙论。我也尝试指出海德格尔对三位前苏格拉底哲学家(巴门尼德、赫拉克利特、阿那克西曼德)的阅读,也是由回到一种宇宙技术,找到另一个开始(anderer Anfang)。
而如果我们认同当前的问题,即科技的全球化所造成的后果,那么现在就是一个机会,很多条件都相应地成熟了。中国的经济和技术发展已到了一个相当的高度。美国不可能再继续原本的单边全球化,即美国输出、我们进口的全球化。我现在要处理的问题是这种新的技术哲学如何发展,我们需要在地地去处理。譬如说大家都在讲智能城市,尤其是中国,目前的智能城市构想其实只是以效率为主,照当前算法和数据主导的思路,伦敦、上海、纽约都会变成一样的,比如智能手机去点餐、支付宝去付款、机器人去服务,这您也可以参考一下电影《攻壳机动队》的虚拟现实和都市主义,这将是我们的未来,而不smart的人将无容身之地,这就是我所说的一种错误的全球化发展。对我而言,智能城市给了我们一个很在地地去重新思考和建构的机会,这也是我最近着手处理的项目。
澎湃新闻:《攻壳机动队》里那样的虚拟现实技术会加深人与人之间的不平等吧?
许煜:那是一种新达尔文主义,是一种人工选择,达尔文主义是自然选择,而现在是人工选择,技术超过了进化。原本 “顺风耳”、“千里眼”是进化的结果,但现在技术改变了人的生理机能。所以,这种人工选择必然是不平等的,比如想让一个小孩变得更聪明进而统治其他小孩,就让他吃药或者注入一个芯片到他的大脑里。我们必须要正视这个问题,要抗拒超人类主义(transhumanism),因为超人类主义本身就是一种更刺激的消费主义。这种消费不止是花钱买东西获得快感,也是彻底改变你的身体,改变你和世界的关系,譬如你的感知性。所以我们要重新提出一种技术发展的方向,超人类主义对很多人来说都是一种引诱,因为它的卖点是增长寿命,改善情商和智商。想象一下,你可以不生病,可以每天都开心或者拥有一个对您百依百顺的机器人女友或男友,这不好吗?可是,这些并不是解决问题的方法,而只是将问题以新的方式提出来。
在国外,许多人都在谈大中华未来主义(sinofuturism)或者说亚洲未来主义(asian futurism)。在美国每有一个新发明新发现就有一个伦理小组来判断应不应该禁止,但在中国就没有这些限制,所以大中华未来主义就是一种加速主义,我们可以想象,在未来,中国或亚洲的技术加速会比美国更快,这种不平等会更严重。超人类主义者并非政治冷漠的,他们有的人也在读牟宗三,认为机器发展可以达到“智的直觉”,这就是所谓的智能爆炸或者奇点,政治就可以消失,因为人工智能能解决我们解决不了的规划问题。但超人类主义以及大中华未来主义都是二十一世纪的虚无主义,这并非我想要走的路线,相反是我所批判的。我提出的是将感知性带进我们的对科技的理解从而来扭转其发展方向,这也是为什么牟宗三对我来说是个重要的对话对象。
澎湃新闻:宇宙技术是外部,那内部具体指什么?
许煜:内部也就是我所说的认识论的问题,也是因此我们必须系统性地研读巴什拉、康吉莱姆、西蒙栋等。举个例子,发展一个软件,比如微信,遇到的第一个问题便是如何定义使用者之间的社会关系。我和你的关系是两点一条线,在Facebook、在微信上也一样,每个人所在的这个点就是社会原子,而社会则是社会原子的集合;原子之间的关系可以用线和点再现。这是心理学家雅各布·莫雷诺(Jacob Moreno)在上世纪初所指出来的,也被认为是现在社交网络的社会学理论基础。但是,我们也要质疑这种认识论以及它如何被吸纳为消费主义的工具。如果按西蒙栋《在形式和信息概念下重新思考个体化 (L’individuation à la lumière des notions de forme et d’information)》所发展出来的理论(我也相信是很重要的),个体化必然同时是心理和集体的,个体和集体无法分开。我几年前和斯蒂格勒有个发展另类于Facebook的社交网络的研究计划,我们提出以集体而不是个体为基础来发展另类的社交网络,以合作和贡献来取代个人主义,这就是从内部来改变的,也是我第一本书中《论数码物的存在》中谈到的内容,而第二本就是从外部来思考这个问题了。
澎湃新闻:这种内外是您自己的说法,还是某种传统的说法?
许煜:这是我自己发展出来的一套论述,这个论述受近几年来在欧洲出现的所谓的人类学的 “存有论转向”(ontological turn)影响,从某种意义上来说也可以说是对Philippe Descola的自然人类学以及André Leroi-Gourhan的技术人类学的响应,同时也是对于海德格尔的批判。一些人类学家例如 Descola强调多元的存有论,他认为在西方的现代性中自然和文化是对立的,他称之为自然主义,但在其他的文化里头,自然并不是处于这种位置的,比如说模拟主义、图腾主义、万物有灵主义等。这按我自己的语言来说,也就是西方现代的宇宙技术(其实也就是自然主义)已经看不到外部的,因为宇宙已经是一个技术系统了,只能从内部去改造。这是为什么我强调技术系统的内部性和外部性。
澎湃新闻:这个项目的进展如何?
许煜:这个项目持续了很久了,从我十多年前在欧洲就开始,我的研究重点一直是技术哲学、认识论、数码研究等,我在学院的框架里跟一些科技实验室和艺术机构合作过,但我是近两年才跟国内有比较多的接触。我从去年开始到中国美术学院上课,第一个课程讲的是《论中国技术问题》中的某些章节,今年的第二个课程是介绍西蒙栋的技术思想,以及宇宙技术的概念,西蒙栋的《论技术物的存在方式》的第三部分对我发展出宇宙技术这个概念有很深的影响。我希望可以将这些问题提出来跟建筑师、艺术家、工程师一起重新思考,这些题目都非常新,特别是西蒙栋的书还没有中、英文的翻译,但学生们都很积极地将上课内容记录成文章放到网上,这将需要时间来消化。
澎湃新闻:在中国美院会搞一个团队来做吗?
许煜:高士明老师大概一年半前邀请我参与建立网络社会研究所,作为客座研究员,我也只是从旁辅助。高老师在这方面很有前瞻性,跨媒学院的院长管怀斌老师也很支持,这对我是很鼓舞的。但您说得很对,有一个团队专门来处理这些问题是很重要,我和斯蒂格勒刚提出在另一个框架内成立一个“数码研究网络”,高老师也很支持。这个网络是跨学科的,而且是国际化的。我们需要培养一批研究员来探讨这些问题,中国当前主要缺乏的不是资源,而是研究人材和研究方法。学术研究必须是认真严谨的,而且必须有方法,要不然就欲速而不达了。我希望这个数码研究与我们在欧洲所建立的相比会更着眼于中国和中国文化。
另外,一些美国、欧洲的思想家,例如拉图尔前几天也跟我谈合作,我们希望以中国和欧洲的新一轮深入的对话来回应目前的困境,提出一个新的框架来处理现代性的问题。这对我来说是相当紧急的事情,但需要一些资金和人手来运营,我现在还没想到怎么弄。
澎湃新闻:中国之外的其他第三世界国家呢?
许煜:我觉得第三世界国家都需要发展宇宙技术论。从人类学的角度来看,每个文化都有自己的宇宙技术论,需要从自己的文化出发来来重新“现代化”。我们没办法避免技术发展的问题,但不能对技术缺乏深入的反思。非洲也好,印度也好,南美也好,我现在开始和这些地区的学者来一起讨论宇宙技术的问题。亚非拉的团结只是将现在的权力分布重新调节,并没有从根本上去处理全球化的问题,只是从原本由美国、欧洲主导的权力版图进行调整,在贸易、国防、文化交流上进行合作,但如果仍然在延续科技的全球化,它还是无法逃出殖民主义。换句话说,不处理好技术问题,只能是殖民的延续,他们改变的只能是权力的版图,但现代化的进程和内涵不会有变化。
澎湃新闻:您对欧洲或者其他地方的社会运动还有在关注吗?
许煜:我一直都有在关注,但近年来我在做的是建立新的框架。我们反资本主义,反新自由主义,读马克思,辩论认知资本主义,去游行去抗争,现在这种抗争的无力感越来越强,很多人要么就变得犬儒,要么将社会运动变成展览。去年这个时候法国的“黑夜站立”运动,几个月的运动突然之间完全消失,原因之一是刚好到了夏天,大家都需要休假。这个理由当然不是全部,但也不是不重要的,我们可以想象那种无力感。很多左翼的思想家、运动家都对此有反思,这也是为什么这几年,左翼加速主义在学术界引起了不少的讨论,因为他们重新提出了马克思一直被忽略了的技术观,如何通过自动化来将剩余劳动扬弃为自由时间;而被视为极左的法国的隐形委员会也在二零一四年底出版的《致我们的朋友》中的一章 “Fuck off Google”里也指出要发展有伦理的技术来响应这个以算法和大数据作为治理术的时代,他们分开有伦理的技术(technique)和资本主义的科技(technologie)。
但我们也要小心不要陷入一种普罗米修斯主义,或者新殖民主义的陷阱,加速主义就是这样,这就是为什么我要从宇宙技术的角度来重新思考资本、自然与文化等范筹,以及技术认识论的问题。我在上面提到的发展一个有别于Facebook的社交媒体(我们当然也可以想微信)的研究,对我来说更为重要,因为如果我们要对抗平台资本主义的话,这是有效的途径,而且无论在教育上还是实践上都会有长远的影响。我在过去多年的研究中提出了一些理论,得出了一些实践的经验,接下来就是怎么落地继续发展,好像您所说的,确实很需要一个研究团队来系统性、针对性地处理这些问题。
澎湃新闻:我看您之前有写到,科学家不一定要在商业化大公司工作,他们可以自己发明一些技术?
许煜:大家常有一个想法,除非你在大科技公司工作,不然是做不出东西来的。还有另外一种说法,做社会运动的都很反抗技术,他们认为工程师、科学家都是站在他们对立面的。这种想法是错误的。您一直往后退,最后变成一群人一起喝酒聊天,大家对各自的要求越来越严格,然后压抑嗑药等问题接踵而来,这是激进主义者氛围里常出现的现象。再不然就是将黑客和工程师对立,黑客是好的,工程师是坏的,但这也是错误的。黑客这个文化范筹首先就必须被批判的。以前说每个人都是艺术家,现在说每个人都是黑客,听起来很酷,但又变成了消费的一种。我的想法是,如果作为一种改变的力量,不应该这么快就将事情对立。科学家也不要以为所有的社会进程都要以工业化的形式出现。很多黑客运动,开源软件的发展,都是工程师的贡献,其实是越来越多了。我们必须以集体作为中心,作为基本元素发展,这才是我说的网络社会的前进。今天跟您说的这些,都需要扎实的研究,可能需要几代人的时间,但这是必要的,希望未来中国能推动这些研究。

Saturday, November 25, 2017

什麼人訪問什麼人﹕我們需要什麼技術知識,抗爭場地在哪裡? ——專訪港產哲學家許煜


Edwin Lo十月 2, 2017訪問 文章導航
PreviousNext

轉載自: https://m.mingpao.com/pns/dailynews/web_tc/article/20171001/s00005/1506794007940

2011年10月,在OC(Occupy Central,現在被媒體稱為佔領匯豐銀行)認識許煜。我在較早時期於營地短暫教授結他與非洲鼓,帶中學生到現場交流;後來他講禮物經濟時我已經不常出現,中間沒有連繫上來。過了一段時間,拜讀他的書評集《閱讀抗爭》、《佔領論——從巴黎公社到佔領中環》才整理出更清楚的輪廓。 2013年,他共同主理的文化組織DOXA舉辦名為《後佔領:藝術、仕紳化、內戰》為期兩天的亞洲藝術/文化工業以及仕紳化的工作坊,邀請我講述觀塘工廠區的文藝生態。印像中其中一名(頗為挺資的)港大研究員發言後,許煜現場評價他「只說了一堆形容詞,根本沒有批判」,場面甚為尷尬。對我來說他不是一個「為搞而搞的學術人」,唔玩畀面遊戲,是真正在發問的哲學家。

2016年,我入讀倫敦大學金匠學院做研究生,算是他學弟;操德、法語的他則在德國呂訥堡大學任教。香港作為一個時差六小時的地方,香港時事,大家倒是追得很緊。我們的朋友梁穎禮因新界東北案​​入獄,他寫信安慰道:「我老師(Bernard Stiegler)打劫銀行,坐了五年監,在監獄裡讀完哲學學位,後來成為知名哲學家。珍惜監獄的安靜,多看書,多思考。」東北十三子重判前兩晚,我發夢自己與一大班朋友坐監,在裡面作息定時,狂操體能,練到好像漫畫《北斗之拳》的人物般大隻,囚衣都谷爆曬。我們除了遠距離觀看香港腐爛而唏噓,還有零距離在歐洲感受右翼復甦而憂慮。許煜一頭栽進技術哲學世界,目的就是要發明新方法,重思全球化。

出版了《論數碼物的存在》與《論中國的技術問題》後,許煜成了炙手可熱的哲學新貴,飛來飛去演講,訪問不絕。編輯留意到他開始在西方哲學界走紅,找我約談;我說只能夠問簡單問題,複雜的,就不知他在說什麼。因為認識,雖然不是哲學訪談,他也一口答應。剛從俄羅斯演講回來,便相約在Skype作這次什麼人訪問什麼人。

■ 問﹕黃津珏 ■ 答﹕許煜

問:適逢佔中三週年,有沒有新的想法沉澱下來?

答:我認識你也是因為社會運動,但你知道,這幾年來我沒有積極參與。佔中期間與格雷伯(David Graeber)在法國《世界報》合刊了一篇文章,講述佔中和香港新自由資本主義歷史的關係,文章被翻譯成十多種語言。結果有些人不高興,明明是民主運動,你批評什麼新自由主義?關注點都被你扭曲了。問題的因果總需要弄清楚,但民粹主義無論左還是右都拒絕這樣做。舉例說,你要自由,又不會想資本主義有問題;你想反資本主義的,又不會從政治學的角度看問題。匯豐佔領對我的影響很大,那幾乎同時是希望也是絕望。在2000年前後,反全球化還是個首要議題,2005年在香港的反WTO很多人都看到,現在還有誰在談反全球化? 2011年的佔領華爾街,2016年的黑夜站立,好像曇花一現,最後剩下的是天台農場、剩食回收。匯豐佔領這種形式的反抗不是不重要,但有它的局限。可能我們需要另一種方法去理解資本主義和現在不斷在轉型的社會關係,這也是為什麼我相信人文學科仍是重要的。

問:我現在靠的是facebook上朋友發或轉發的資訊去理解現在的香港。你又怎樣?現在你與香港的關係如何?

答:現在facebook也不是經常上,香港報章也不是經常看。近年多去中國大陸講課和參與活動,反而從另一個側面看香港。無奈的是,香港其實有很多優勢,很多基礎設施都已經發展好了,但這種優勢似乎被一種保守主義逐漸地磨蝕。我這樣說可能會挨罵,但你確實可以見到中國大陸有很多實驗的機會,尊不尊重知識是個別的事,但尊重人才方面整體上是比較好。以前有人會覺得,香港是前殖民地,中西合璧,視野應該較廣闊,現在聽到也會臉紅,中西合璧的例子你也只能舉餐蛋麵和鴛鴦吧。實驗很重要,沒有實驗就沒有新的東西出現,新的文化與文化進展都是來自實驗和對歷史的反思。

單邊全球化與反動主義

問:之前在倫敦與你談到香港引發的右傾現象,源自「被扭曲的全球化」,可以解釋一下嗎?

答:我嘗試論證到目前為止,全球化一直都是單邊的全球化,它主要是基於技術上的差異(就好像兩次鴉片戰爭之後中國被迫現代化)。技術上的差異即是有一方比較先進,他們輸出產品,產品附帶著許多意識形態與社會規範。但你發現這個技術上的差異是不斷被消除的,在消除的過程中,輸出方面的優勢就會逐漸被弱化,那樣就會產生反動意識。現在美國的新反動運動(neo-reactionaries)以及另類右派運動(alt-right)都是這種單邊全球化走到盡頭的結果(這跟德國威瑪時期的反動思想的因由是不同的), make America great again;Brexit就是make Great Britain great again。 Donald Trump的例子更清楚,因為他不斷指摘中國拿了美國很多著數。而在Trump的競選過程中,他聲言要把在中國的工作帶回美國,怎樣做到呢?就是在美國境內實行自動化生產。因為中國的自動化已經發展到一個位置,就是美國絕對的技術優勢已經動搖了。如果我們將美國與中國的關係,放回香港與中國的關係,是一樣的。這些反動主義的理論基礎是基於拒絕啟蒙運動的普世價值,但他們就好像黑格爾在《精神現象學》裡所說的「苦惱意識」(unhappy consciousness)一樣,意識到矛盾又無法超越。我們見到,右翼在拒絕普世價值,左翼在肯定普世價值,被人罵成「左膠」。我認為必須超越這種簡單的二元對立,甚至比右翼走得更遠,也就是從肯定差異而發展出共同;但最關建的問題是如何切入構思另一種全球化,這對我來說是技術的問題。

想怎樣轉化現有技術?

問:早前Pirate Bay創辦人Peter Sunde說互聯網失敗了,現在能做的只是減低破壞,你有什麼想法?

答:這個說法完全錯誤。互聯網在九十年代普及時很多人把它看成烏托邦,尤其是「安那其」,將它看成是沒有階級秩序的空間。近年來,隨平台資本主義(如facebook、Uber、Airbnb等)的出現,「敵托邦」已成無可否認的事實。但重點不在於如何破壞以及反監視,而是怎樣把技術重新居有和轉化。何謂將技術重新居有和轉化?我舉facebook作例。任何一個軟件、一個平台都有它自身的認識論(Epistemology)。有一種認識論,也就會有其他的。 Facebook的認識論是怎樣的?它怎樣讓你認識社會關係?就是畫兩點中間穿一條線。社會心理學家莫雷洛(Jacob Moreno)早在上世紀三十年代就是這樣理解社會網絡,「集體」(collective)就是很多點連著很多線。但你知道「集體」不止是這個意思,人類社會不一定是由社會原子(social atoms)開始,不一定是N個個體然後他們有了以物易物系統才誕生「集體」。莫斯(Marcel Mauss)說禮物經濟(Gift Economy)時,一開始便指出,這種個體之間的交易是「經濟的最初狀態」其實是個假象。因此問題就是在facebook提供了一種認識論的時候,怎樣去用現在掌握的技術去發展其他認識論。其實可以有很多這樣的嘗試,好像我早幾年在巴黎有一個研究項目,我們發展了一個基於群組而非個體的社交網絡雛形。那是2012年,公眾都不是太意識到這個問題,現在就不同了。好像上星期我在俄羅斯提出了,莫斯科的一個知識分子媒體又來訪問,現在外界有很熱烈的迴響。因為以前大家都假設了它的必然性,人的關係就是兩點一條線,但這些都必須被質問。所以,我們不止是要去減低破壞,而是要想怎樣去轉化現有的技術,這個很重要。

另外,我們也需要回到什麼是技術這個問題,這也是我在《論中國的技術問題》一書對海德格爾的技術哲學的回應。這本書的德文版和中文版會在明年出版,我期待更多的討論。如果我們看現有的技術,好像技術是普世性概念:無論你是中國人也好,希臘人也好,用的都是同一套。我也聽過一些中國哲學家用普羅米修斯(希臘神話)來解釋中國技術的源頭,那是完全的失向(dis-orientation)。我嘗試指出的,是這裡有一個康德的二律背反(Kantian antinomy),就是「技術是普世的」論題,因為技術是記憶的外化(exteriorization of memory),器官的解放(liberation of organs),人類學而言這個就是成為人類的過程。而它的反論題就是「技術不是普世的」,每種技術背後有它特定的宇宙論作為它的知識型,例如我們沒辦法用古希臘的technē去解釋中國的技術。裡面不可化簡的是什麼?它們可能出於偶然,但在歷史過程中成為必然。如果技術之間存在著不同化簡性,那各種技術都有著不同的系統性的思考,我稱之為這些為宇宙技術(Cosmotechnics)。我希望非歐洲文化可以通過梳理這些技術思想來重新思考對現代技術的居有,克服一個單一化的技術未來。在這種單一的未來的想像裡,地緣政治只是在競爭誰早點到達技術奇點,像《人類大命運》描寫的荒誕的「神人」(homo deus),背後其實是矽谷意識形態:超人類主義、人工智能、深度學習、智能城市……最後是徹底的去政治化,因為超級智能可以做所有的規劃。

中醫的宇宙技術

問:我有讀過你的《論中國的技術問題》,第一樣想到的是中醫與西醫的問題。

答:我用中國作為例子來解釋宇宙技術這個概念對克服現代性的意義,但不止是中國,所有的文化都可以而且必須重新發現它們的宇宙技術,我現在跟印度以及南美的學者討論這個問題。中醫是個好例子,好像古中國的宇宙論,中醫也會用到一些術語,如「陰陽」,如「氣」、「五行」的概念。所以中醫是一個很重要資源去分析我所說的中國的宇宙技術。西醫基於不同的認識論,與中醫的不一定相容,兩者也不可化簡,就算古希臘醫學也有把脈,但跟中醫的相去甚遠。現代化是怎樣的過程?其實是同步化過程,即是所有事情必須化簡成同質性的規格。你去化驗中藥裡面的化學成分,然後與西藥比較,就是同步化。如果我們想避免單一的技術未來,同時拒絕一種黑格爾式的世界歷史,我們必須要重新處理技術歷史的問題,想像可分支(bifurcate)的未來,以另一個框架來重新居有現代科技,而不止是否定它。當然這需要很漫長的努力,可能要幾代人才能完成。
連結運動創造新知識

問:現在東亞地方開始建立自己的安那其連結,第二屆名為No Limit的藝術行動剛結束,日本哲學家柄谷行人也在這個連結中活躍起來。之前與你傾談時提到一些建議,能否再講述一次?

答:我沒有參加,不知道那具體是什麼,不便評論。我只能說我自己的想法,對我而言連結運動最重要的是知識的生產。我們必須在連結過程中創造新的知識,而不止是一起打邊爐飲啤酒,你記得我們的朋友日本社會學家樋口拓朗雖然也是素人之亂的一分子但他很批判快樂抗爭。我們四年前舉辦士紳化的討論會,無非也是想要處理藝術和抗爭之間的辯證性關係。馬克思與恩格斯在《德意志意識形態》裡面提出「自由人」(freeman)的說法,什麼是「自由人」?大概就是說你早上釣魚,下午打獵,晚上可以辯論時政、哲學。柄谷行人因為馬克思在《法蘭西內戰》以及《資本論》提出了自由人的聯合,所以說他也是安那其,但他沒有深入探討馬克思說的「自由人」是什麼。馬克思自己在《政治經濟學批判》寫到工廠的自動化,減少了必要勞動,卻增加了剩餘勞動,他提出的辯證是如何把剩餘勞動揚棄(sublate)為「自由時間」(free time)。什麼是「自由時間」?馬克思說對他而言「自由時間」不是傅立葉(Charles Fourier)說的遊戲玩耍,而是休憩時間與高等活動(higher activities)。什麼是高等活動?就是科學、技術、人文等知識的生產。不是生產一種被邊緣化的知識,而是生產一個可系統化的知識,或者將被邊緣化的知識轉化為可居有的知識,它能容許我們去回應目前出現的問題。現在很多人在辯論,說自動化生產已經出現,我們進入了後資本主義,但我們怎樣理解馬克思百多年前說的話?自動化生產是否必然衍生出「自由時間」?當然不是。在2013年起開始在歐美流行的加速主義(Accelerationism)提倡左翼不要漠視技術,不要滿足於天台花園等無法規模化的抗爭,而是重新利用現有技術,加速到完全自動化,就有可能瓦解資本主義獲得「自由時間」,前者不是沒有道理,後者是天真的。這個盲點我們一定要清晰理解,思考21世紀我們需要什麼技術知識,抗爭的場地在哪裡。

■答﹕許煜,哲學家。任教於德國呂訥堡大學哲學研究所與數碼媒介美學與文化研究所,曾在倫敦大學金匠學院及巴黎龐畢度中心創新研究所工作與授課。

■問﹕黃津珏,音樂家,藝術家,文化評論人。倫敦大學金匠學院音樂系博士生。

文﹕黃津珏

編輯﹕蔡曉彤

许煜:数码化时代科技和人文的契机 Edwin Lo十一月 18, 2017

許煜德國呂納堡大學哲學研究所/ 中國美術學院

[導讀] 21世紀以來,人類社會科技的發展進入一個新的井噴期,不同於科技一日千里的發展,人文領域的思考範式卻停滯不前,社會想像力日漸枯萎,事實上,現代科學的發展有著深厚的宗教、哲學、文化等思想根源,早期科學的發展與西方人文主義的演進一直是交互影響,二者之間有著緊密的聯繫。但在隨後的歷史進程中,科學領域與人文領域逐漸分離。正如許煜所指出的,在我們思考人文危機的時候,問題早已不在人文,而是整個知識體系的分工及其結合將會面臨巨大的挑戰,由此,新秩序的出現也成為可能。這要求我們回到自身傳統的知識脈絡,來思考面對數字化和信息化的進一步發展,將有怎樣的選項和可能性留給中國。科技危機或人文危機?科技與人文是兩個巨大無比的範疇。編輯在約稿信中提到“科技發展與人文衰敗之間所形成的張力,構成了現代性危機的重要表徵之一”,但是,人文衰敗到底是因為國內對於人文學科過於悲觀,對西方的人文一直對技術問題的反思視而無睹,還是對科技過於依賴?這到底是人文危機,還是科技危機?是人文追不上科技,還是科技脫離了人文?如果說是科技加速導致了這樣的一個危機,那意味著科技和人文之間不但存在著張力,而且還存在著鴻溝。就好像科技和人文之間一直存在著間隙,而從人類人明進化的某一刻開始,這個間隙突然之間被拉大了,似乎出現了巨在的裂縫;驟然之間,我們發現再也無法跨越過去。如果今天我們要處理這個危機,就必須疏理清楚,到底這個間隙來自哪裡?而人文學科又是如何處理文明發展中所出現的文化危機?在我們開始進入討論之前,必須理解所謂的人文危機,並不只是出現在人文科目中的危機,而是總體性的文化危機,也是西方現代性的表徵之一。早在上世紀三十年代,胡塞爾在《歐洲科學的危機》一書中針對歐洲科學的技術化,就提出了發展現象學作為科學的根基的看法,也即重新從主體的經驗場域出發;晚期的庫爾特·哥德爾,現代計算機歷史上的重要人物,也在胡塞爾1931年的《笛卡兒式的沉思》當中找到了迴響。而同期,除了歐洲科學的危機外,還有文化危機、數學危機 、物理危機、機械危機等等。如果說上世紀的人文學科還想重新成為科學和技術的基礎,在今天我們卻幾乎很少看到這樣的嘗試,相反,我們見到的是人文學科看到了自己的局限,同時又無法克服這種矛盾。而到底科技對人文的超前(如果我們可能用這個字的話)是怎樣出現的?二十世紀的人文學科對技術的反思又是如何進行的呢?在闡釋觀點之前,筆者必須承認我們將不得不很概括性地來進行這個工作,而因為文章的篇幅所限難免會出現疏漏,而筆者也相信在不同學科里都已有一些學者在努力地進行新的反思,這篇文章只是想歷史性地討論人文危機,遠非要貶低這些在進行中的工作。與此同時,我們也要從中國文化的觀點來思考西方人文學科所進行的內部反思,這個思考必然是詮釋學式的,也就是說,它必須將西方人文學科的歷史進程置於中國自身的歷史進程來思考。這篇文章嘗試先分析西方人文學科對於現代科技的應對,進而來理解中國語境下所出現的“人文危機”。技術問題的壓抑 西方人文學科一直都在壓抑技術的問題,這裡所說的壓抑是弗洛伊德意義上的Verdrängung。這個說法,早已由幾位西方哲學家提出,最為人知的是法國哲學家貝爾納‧斯蒂格勒,而早於斯蒂格勒之前,另一名法國哲學家佐治‧康吉萊姆提出了西方哲學一直都在迴避技術問題(除了一些原子論主義者之外),直到笛卡兒的著作裡,技術才成為哲學的基礎。但技術在笛卡兒的哲學裡指的是機械論。笛卡爾在《談談方法》中提出一個廣為人知的想法,他說,當他看到窗外的街道上走的行人,他懷疑這些到底是不是穿著衣服、戴著帽子的機器人。在他的其它著作如《論人》以及《人體的敘述》中,他將人當成了機器來形容,其中他提到的一個例子就是教堂裡的管風琴,氣就好像是靈魂或者動物精氣,而管道就好像是血管一樣。然而,笛卡兒的機械論並沒有站得住腳,因為與機械論對立的是自由的概念,也就是說機械論在威脅著自由,在一個完全機械化的系統裡,將沒有自由可言(這也就是康德在《純粹理性的批判》裡的第三個二律背反提出的問題,而他在之後的《實踐理性的批判》、《判斷力的批判》繼續探討的問題)。這也是為什麼後康德主義者,特別是謝林和黑格爾的思想一開始就是從高處來把握以及消化掉機械論。相對於康吉萊姆視笛卡爾為首位系統性地將技術置於哲學核心的哲學家,斯蒂格勒則指出早在柏拉圖的對話錄裡,技術問題就是核心性的,因為如果沒有技術,也就沒有回憶,也就是說沒有真理;就好像在《美諾篇》中,奴隸之所以能夠抵達真理(解答幾何問題),是需要通過回憶的。要不然的話,我們便無法響應美諾對蘇格拉底提出的難題﹕如果你一早知道真理是什麼的話,那你就不用找了;而如果你不知道真理的話,就算給你遇上你,你也認不出它來。蘇格拉底的回答是,他一早已知道真理,只是他遺忘了,因為靈魂在每次轉世的過程中都會忘記前世所發生的,但忘了的東西可以重新記起,如奴隸依靠的是通過沙上畫下的痕跡來“回憶”如何解決幾何問題。技術問題作為形而上學(如果我們跟隨海德格爾所說的形而上學由柏拉圖開始,而前蘇格拉底的哲學是前形而上學或者非形而上學的話)的核心問題在西方哲學中是被長期的邊緣化處理。技術哲學,嚴格來說從十九世紀才開始出現,也就是說,它作為反身性的哲學開始真正意識到一些東西正在超越它所是,並嘗試將其內化為它的系統的一部分。早期的學者如德國哲學家、黑格爾主義者恩斯特·卡普在1877年出版的《技術哲學的基本路線﹕從新觀點理解文化史》就開始將技術視為人類器官的投射。但是,為什麼哲學在十九世紀才開始意識到技術這一不斷膨脹甚至超乎其自身的存在呢?也就是說,哲學是怎樣從現代性的技術非意識進入到一種技術意識的呢?最根本的原因,在於十八世紀末出現了工業革命,科技形成了一股巨大大的力量,這也是為什麼海德格爾在一九四九年的講座《論技術問題》中指出了現化科技的本質不再是古希臘詩意的technē,而是將一切都視為可剝削的原料的座架。也就是從十八世紀末開始,歐洲科技的形像出現了轉型,如另一位法國技術哲學家吉爾伯特·西蒙東所說的,科技變成了強暴自然母親的暴徒。因為如此,對技術的仇恨也開始出現了,而在馬克思描繪的工廠勞動中,機器成為了異化的工具。在海德格爾發表了他對現代科技的批判的同時,在美國出現了“控制論”,而其后海德格爾更聲稱控制論的出現標誌著西方Metaphysics的終結。控制論是現代西方一個相當重要的運動,這個運動的重要之處並不只在推進了所謂的“自動機理論”,而是他的奠基人諾貝特‧維納發展出了“反饋”這個概念及其在數學上的應用;與維納的控制論同期出現的,還有約翰·馮·諾伊曼的自動機理論,卡爾·路德維希·馮·貝塔郎菲的一般系統理論,這些學科對於我們現在所身處的“控制論化” 社會有非常深遠的影響;維納的控制論還有一個更為重要的影響是,他想發展出一個機制來連結以及統一所有的學科,他在《控制論》中提到,辦公室都在同一走廊的同事,就算是同一個學科,也因為學科內部的分工,而無法一起進行學術交流;他希望用控制論的反饋機制來促進各學科的交流和合作,這也是為什麼傳奇式的每年一次的梅西會議會集合數學家、物理學家、社會學家、語言學家等進行跨學科的討論和合作(當時也有中國的參與者,語言學家趙元任)。如西蒙東在一九六一年發表的一篇文章《技術思想》中指出,西方歷史上的兩種技術思想或技術精神,第一是笛卡爾的機械論,其二就是維納提出的“反饋論”,前者是基於線性的機械因果,後者容許非線性的邏輯和互動。也就是說,在笛卡爾和控制論之間出現了兩種非常不同的技術思維,而這兩種技術思維的目的都是要成為所有學科的基礎,其後,我們也看到第二代的控制論,包括海因茨·馮·福爾斯特、弗朗西斯克·瓦雷拉(我們甚至可以將德國社會學家尼克拉斯·魯曼、法國社會學家埃德加·莫蘭等也算進來)等在維諾之後更進一步發展其理論;而在上世紀七十年代,控制論甚至成為了智利社會主義經濟改革計劃的核心思想。歷史與詮釋 以上歷史描述主要是想指出,在歐洲哲學和技術之間的互動,雖然如斯蒂格勒所說存在著一種壓抑,但同時也存在著一種連貫性。在中國(以及在大部份非歐洲的文化)這個過程則是完全不同的,因為科技是從外輸入的。從鴉片戰爭之後,中國被迫使採納西方的科學、技術和教育制度,這也是我們現在所說的中國現代的開始(粗略來說改革開放以前,中國現代化有三個時期:洋務運動、五四運動和文化大革命)。當然,我們也必須承認科技位於馬克思主義思想的核心,因為它對於人化相當重要。這一點恩格斯在《自然辯證法》已經清楚表達過了,而且後來也成為國內馬克思主義的核心科學觀。一直到1990年代,中國都沒有的科學和技術研究和技術哲學之類的學科,它們全都被放在 “自然辯證法”之內 – 這是恩格斯一份手稿的標題。然而,這份手稿的其中一章《勞動在從猿到人的轉變過程中的作用》用人類學的方式解讀技術物在進化過程中的重要性,但同時它也假設了一個普遍的技術概念。然而,前述的哲學和現代科技的互動歷史在中國則是缺席的,而這種缺席,也正是我們上文提出的詮釋學理解的條件。這段歷史的缺席的直接後果,就是沒有足夠的能力去反思它,因為在中國的哲學語言裡缺乏可以理解它的概念和範疇,最後也造成了一系列錯譯,如技術哲學家李三虎先生所描述的道器關係的顛倒,而拙作《論中國的技術問題》則以知識型為核心來重構這段歷史;但也因為這個缺席,中國可以像美國一樣加速發展技術,甚至超越美國,因為在中國,技術發展是幾乎沒有阻力的;在美國,當一種新的技術出現的時候(如克隆)就會有一個倫理小組來製衡它,但這在中國卻是另一個故事。因為這個特殊的語境,我們必須謹慎地思考這樣一個“人文危機”,而不只是一窩蜂地去追逐數字人文、網絡文化等在西方已發展起來的學科。在回到我們的主題——人文危機和中國技術問題——之前,我們先概覽一下西方人文科學是如何系統性地處理技術問題的。讓我們留意一下前文所說的“壓抑”一詞,如果有什麼東西是被壓抑的話,這也表示它其實一直都是存在的,也就是說,技術問題一直都處於人文學科之中,只是被掩蓋了以及忽視了而已;換言之,它必須被重新辨認出來。要處理這個危機的話,我們必須將技術問題從人文學科中重新辨識出來、,而哲學則是其中最為古老的學科。這也是為什麼,斯蒂格勒的著作對我們而言是非常重要的,因為他不只發展出一種獨特視角的技術哲學,而且對西方哲學史作出了一種新的解讀,而這種新的解讀也意味著重塑與力量。當然,人文學科並不只是哲學,還包括文學、社會學、人類學、語言學等。西方除了有歷史積澱的技術哲學之外(包括歐陸技術哲學,分析技術哲學,後現象學技術哲學等等),也還有其他研究技術的人文科目,例如﹕媒體研究、科學和技術研究,以及由其洐生出來的媒體考古學、文化技術、媒體哲學等等。姑且不論與計算機科學緊密結合的語言學,文學也因為它的靈活性,可以說是最早對數字技術有系統性反思的學科,上世紀有關數字文化的重要著作都來自文學系或文學出身的學者,例如美國杜克大學英文系的凱瑟琳∙海勒絲,她1999年出版的《我們是怎樣成為後人類的》至今仍是重要的作品;而我們現在所說的媒體研究以及數字人文都是來自對文獻、檔案、影像、電影的研究;又或者,德國的費德里希‧奇特勒,這名在弗萊堡受教育的學者,是德國媒體理論的奠基人,他的著作《摹記系統1800 /1900》與《留聲機、電影、打字機》是媒體理論的重要讀本,今天德國幾個重要的媒體研究中心,如柏林洪堡大學、魏瑪、呂納堡等大學等都受益於奇特勒。另外還有在歐美幾乎走到盡頭,現在又在別處興起的科學和技術研究,則主要是社會學以及人類學(特別是基於布魯諾·拉圖爾、米切爾‧卡農、約翰‧羅等行動者網絡理論對技術問題的反思。

也就是說,這些學科一直都在反思技術和人文學科的關係,至於說“人文危機”這個字,除了其新聞式的噱頭意義之外,其實也表達了某種理解的不足;但是這種不足,又不是全然沒有道理,因為這些反思確實並不徹底。試舉兩個例子,如在德國的媒體理論開出來的“文化技術”或者“媒體考古學”專治的是媒體歷史的研究,如不少論者指出的,這些研究的精細程度可以跟科學史媲美,但是也如大部份的科學史研究一樣,這些研究將當前的媒體當成死物的研究,對於其政治意義則相當漠視;還有,數字人文基本上可以說是科技對現有學科的研究方法的衝擊(例如弗朗哥·莫萊蒂在2000年提出的遠距離閱讀),或者用計算機程序來分析文本或者畫風(例如列夫·曼諾維奇在2008年左右提出的文化分析學,其中包括用影像來分析蘇聯的形式主義作品之間的關係),而並不是對工業技術的批判;又或者許多做STS的學者,研究的是臉書、微信等造成的社會現象,這並不是說他們做的不好,其中有不少研究相當嚴謹而且有趣,但是問題在於它們都變相地成為了對這些工業媒體的服務”。

這遠非批評這些學科毫無可取之處,這裡面也有非常有貢獻的研究,而是我們需要(1)更徹底地重新思考人文學科和技術的關係,除了針對新技術產生的社會現象的觀察和評論之外,人文學科如何參與技術的進程;與這個相對應的是,未來的教育系統如何去消化這個所謂的“人文危機”,未來學科的分工必須如何發展? (2)中國如何去把握當前的這個決定性時刻,重新以詮釋學式的方法重新佔據這些技術發展,同時避免現代和傳統的進一步割裂?筆者在這裡無法全面回答這些問題,只希望提出一些觀察和想法。
人文與科技結合的新知識系統

讓我們先來看第一個問題,如果我們同意人文學科在處理技術問題上總是繞著工業技術走的話,那麼問題是人文學科如何介入當前的技術發展和批判?如果人文學科的介入只停留在歷史研究、價值批判或者為資本主義造成的破壞提供精神治療的話,那麼所謂的“人文危機”將會繼續下去;而如果人文科學希望能夠介入技術想像的話,那麼這些人文科學必須和工業技術建立一種新的關係,也就是說,它必須參與以及引導工業發展的方向。在國內外很多左翼學者都在批判“平台資本主義”,也就是說工業技術以及資本主義的一種新發展,如臉書、谷歌、微信等,用用戶的數據來轉化成資本以及新的消費模式。這些平台帶來了很多的便利,但是也造成了一種以消費為主的技術發展模式;這並不是說沒有這些平台,我們就無法獲得這些便利,問題在於我們如何能夠發展出另類的模式?

筆者相信人文科學不單是可以參與,而且能夠/必須大力地推進一個新的方向。現代社會的技術知識出產一直都呈現為兩極,一個是擁有豐富知識的工程師階層(包括黑客),另一個是只擁有使用能力的普通人,兩極之間幾乎沒有中間者。筆者認為人文科學必鬚髮展出第三個可能性。筆者在二零零七年開始便參與了一系列以人文學科和科技學科合作的項目,包括倫敦的泰特美術館的視頻檔案庫的試點工程、龐比度中心的創新研究所研發的不同於臉書的社交網絡(這個項目也會與荷蘭的網絡文化研究所合作繼續發展)、以及和普林斯頓大學和德國電信發展的另類推薦系統,前者的靈感來自胡塞爾的現象學,而後兩者分別來自西蒙東的集體個體化以及締合環境。這裡涉及的同時是認識論以及存在論的問題,而這些問題,大部份的工程學院(據我所知除了法國的某些工程學院會涉足之外)都是不會教授的,這些問題不但是批判性的,而且同時打開了技術發展的新的可能。然而,這些需要新的嘗試,包括對專業和經驗的引進和尊重,問題是有沒有學院願意做出這種嘗試?

最後讓我們回到中國的技術問題,而思考這個問題的時候必須先從全球政治出發,再回到當前的要務。工業技術無論是國內還是國外都在走向一種超人類主義,也就是說全面自動化、人的身體、智能、情緒的優化(如我在另處所說的一種高端的消費主義),中國日前推出了《新一代人工智能發展規劃》,確立了中國在二零三零年要成為國際人工智能創新中心的目標;普京在俄羅斯九月一日的“知識日”(開學日)向兒童表示,未來誰在人工智能領先就可以領導世界;也就是說,未來的技術發展,也就是向著奇點的發展。姑且不論這種發展是否可取,但人類文明的愚昧恰好是因為它誤以為是沒有選擇的,你只有去或者不去,而不是想去哪一個方向。數字知識型在現階級已經存在了:只要我們看看媒體是怎樣大肆宣傳數字媒介、創新、人工智能、社交網絡、智慧城市、物聯網等等就明白了。它們構成了一個新的真理體制。數碼科技正在迅速變成文化、經濟、社交等等的基礎。西蒙東六十多年前寫《論技術物的存在模式》時,曾觀察到文化和技術被差別對待,但這個對立在今天卻有了另一個形式:科技成了驅動文化的主要力量,它調節著文化的動態。中國科技可以在中國思想缺席的情況下存在嗎?當然可以,這就是我們現在的狀態,也是我所理解的大中華未來主義一詞的意涵,也是海德格四十年代中期在《黑色筆記本》中預期中國的政治進程時,說到“科技將自由進入中國”的意思。

我嘗試論證技術的多元性,也就是西方技術與非西方文化的技術包括中國技術、印度技術等有著很大的區別;譬如說,我們不能將中國的技術簡化為古希臘的technē或者歐州現代時期出現的科技,正如將西方的Metaphysics當成中國的形而上學一樣,如果海德格爾可以說控制論是西方metaphysics的終結(而這種終結包含了結果的意思),那中國的形而上學則是換了另一種方式終結了自己。這也是我嘗試指出不同的文化中存在著不同的宇宙技術,它們有著非常不同的認識論和知識型。在當前全面邁向奇點的過程中,我們還能夠逆轉當前的形勢開發出不同的技術嗎?

筆者《論中國的技術問題》一書嘗試從這個現代化和科技全球化階段,回到歷史,回到傳統形上學,來理解有怎樣的選項和可能性留給我們。就如讀者知道的,福柯在《詞與物》 之後放棄了“知識型”這個術語,而開始使用dispositif(部署),並將知識型重新定義為一種部署。我改造了知識型一詞,賦予它政治的能動性。筆者所想追問的是,將來有可能想像出一個新的知識型嗎,從而用它來找到另一個框架數字科技的方法?文化陷入危機時,會被迫製造出一個新的知識型,作為新的感受力和新的感受方法。
筆者越來越相信這曾是利奧塔想用後現代這個概念所做的。 2013-2015年期間筆者領導過一個關於利奧塔和夏樸1985年在龐畢度中心所策展的“非物質”與後現代理論的研究計劃,而我直到最近才開始理解,後現代之於利奧塔是另一種知識型,因為它涉及另一種與現代完全不同的感知性。也就是說,繼福柯分析的文藝復興時期、古典時期、現代的三種知識型之後的最後一種。然而,在中國脈絡下,知識型問題必須從中國自己的歷史出發來考察。

筆者想在中國哲學史中區別出三種知識型:一、先秦哲學的出現以及周朝結束後儒學逐漸佔據統治地位,它在天人之間建立了、也正當化了道德感受力,因為天提供正當性給政治行動、社會行動和個人行動;二、晚唐佛學佔據統治地位後,11世紀出現了新儒學,它重建道德宇宙論,重新把宇宙生成學引入儒家思想,來重新肯認宇宙和道德之間的統一;三、中國在鴉片戰爭被英國擊敗後,被迫尋求新的知識型,來處理西方科學和科技,可是卻失敗了,因為中國嚴重缺乏理解科技的知識,也嚴重缺乏掌握這種物質轉型的經驗。對我們而言,現在似乎正是重新認真考慮對認識論和知識型的探究的關鍵時期,因為全球化現在已經走到極限,應對人類紀變得越來越迫切了﹕矽谷的超人類主義者追求著全面的去政治化,生態現代主義提倡用地理工程學來改造地球和生態,這些都是西方現代知識型的延續;而在中國的語境下,這種新的知識型將可以是什麼,可以用它來找到另一個框架和引導數字科技的方法嗎?

希望我們能夠帶著這些疑問去思考人文危機的問題,去理解其中的歷史性以及復雜性,同時去重建人文與技術的關係。不少研究科技的學者都預言,當數字化和信息化進一步發展的話,到時出現的不再只是學科的分裂,而是整個大學制度的瓦解。或許在我們思考人文危機的時候,問題已不在人文,而是整個知識體系的分工及其結合將會出現的巨大的變化。但是,瓦解並不是世界末日,而是另一個可能;這也是為什麼在這個時候我們必須為它的瓦解做好準備,因為瓦解也正是新的秩序出現的可能。
許煜的是媒體歷

Wednesday, November 15, 2017

Interview with Yuk Hui: Digital Objects and Metadata Schemes


Yuk Hui has dared to pull philosophy into the twenty-first century by asking what a digital object is. Originally from Hong Kong, he has been roaming Europe since 2006. He first did his PhD in London at Goldsmiths College, then relocated to Paris and worked at Bernard Stiegler’s Institute of Research and Innovation before moving, inevitably, to Berlin, where he is a postdoc at Leuphana University (Lüneburg). His first book, On the Existence of Digital Objects, arranges a dialogue between the technophobic metaphysics of Martin Heidegger and the French technology thinker Gilbert Simondon (author of the neglected 1958 classic On the Mode of Existence of Technical Objects). In his debut, Yuk Hui elegantly plays with the double meaning of the word “ontologies”: on the one hand, the eternal level of the question of Being ; on the other, the technical meaning of the word used by computer science to describe the hierarchies inside representations of knowledge such as metadata.
Ontology in the context of the internet is often associated with the inventor of the World Wide Web, Tim Berners-Lee, and his term “semantic web,” a set of standards for data formats and exchange protocols. One way to describe On the Existence of Digital Objects is to say that it gives the touching yet superior engineering mindset of Berners-Lee a solid continental European foundation. Programmers do not just hang out on Slashdot, 4Chan, and Reddit; they also read Husserl. Indeed, some hyper humans might … My question is why the geek establishment didn’t foresee the rise of platform capitalism, with monopolies such as Google and Facebook. Information science’s approach to ontology has proven naive, if not shortsighted. The internet as a public realm that the engineering class takes for granted has all but disappeared, leaving no space to implement experimentation on the fundamental (indeed ontological) level. This raises the question of whether ontological adventures such as this one can be successful without a political angle.
According to Yuk Hui, “The idea of the philosopher as a figure who stands outside as mere critic and defends the purity of thought has been washed away in the flux of technological progress.” The nature of technics needs to be taken into account when talking about being. That’s an ambitious starting point. However, the real existing social media dominance puts on the table the question of what role philosophical investigations (such as Hui’s) can play. Should research become more technical (and necessarily more traditional in order to be accepted)? Or should it go against the grain and refuse to build foundations in the service of an insular engineering class that is in dire need of a Žižek-style political provocation? Another approach could be to compare Hui’s surprisingly Deleuze-free style with American programmer-theorists such as Alex Galloway and Wendy Chun, who have never dug as deep into classic philosophy in search of the foundations of our digital existence. Who’s ready to read XML syntax alongside Schelling and turn knowledge of Python and C into action, thereby changing the language of philosophy itself?
At times, On the Existence of Digital Objects falls into the obligatory comparative exercise of explaining how author A is unlike author B—but then it recovers quickly, giving us a sense of things to come. What’s really upsetting about the future of this digital philosophy-in-the-making is the “black box society” (Frank Pasquale), the secretive algorithms that cannot be read, let alone changed. How can philosophy become technical when it, once again, can only speculate about its object?
Let’s praise Yuk Hui for his priceless effort to practice what Friedrich Kittler always proposed, yet towards the end of his life drifted away from, escaping to Ancient Greece. Bernard Stiegler’s preface to Hui’s book is equally appreciative. Next stop for Yuk Hui is a similarly ambitious study on the nature of technology in China, which he has just finished. Let’s now get to the subject: the digital objects that surround us, and steer us, in such virtual, invisible, and intimate ways.
Geert Lovink: Can you sketch the long-term implications of your approach for philosophy at large and how it is taught? Where are we in terms of the debates and experiments to integrate technics into the philosophy curriculum? Networks and philosophy have yet to encounter one another. How do you want to stage this? Some say that the “encounter” is a Christian notion to start with.
Yuk Hui: Like Bernard Stiegler, I am trying to reread philosophy according to the question of technics, not only within European philosophy but also Chinese philosophy—for the latter I am collaborating with some Chinese scholars, for example Professor Gao Shiming from the China Academy of Art. Stiegler is a very good example of this since he bases his reading of the history of philosophy on what he calls the “tertiary retention,” which is artificial memory. Tertiary retention is a supplement to what Edmund Husserl calls “primary retention” (impression) and “secondary retention” (recollection). Stiegler develops his reading in a systematic and rigorous way. However, we still need to do an enormous amount of work to take this further, and necessarily with a “collective” if not a school (and indeed Bernard has a philosophy school in Épineuil-le-Fleuriel), which will firstly have to deeply engage with philosophical texts and the philosophical tradition instead of mere intuition, which is always necessary but not sufficient; secondly, it will have to closely engage with technological development, and in this regard it is necessary to work with engineers; and thirdly, it will have to take the concept of technics beyond Western discourse, which seems to me a very urgent task in the Anthropocene.
You said that networks and philosophy have yet to encounter one another. I would say that such encounters are immanent. We can always see the question of networks in different thinkers, implicitly or explicitly. For example, it’s clearly evident in Saint Simon, Marx, Heidegger, Simondon, Deleuze, etc., not to mention in more contemporary philosophers; however, we need to retrieve and thematize these thinkers—“in the Christian sense,” as you said, like the encounters of Christ in the Gospels—in order to respond to the problems of our epoch. This is exactly the point I have made before.
GL: What went wrong with the corporate discourse around Big Data? What’s so boring and suspicious about it? And why haven’t the “digital humanities” risen up against this monstrosity? Would you be in favor of data being discredited altogether? Or would you rather say: another data is possible? Recently, a “data prevention manifesto” was posted on the nettime list. It argued against protection and the “privacy” paradigm. We would be much better off, it said, preventing the production of data in the first place. Would you say that data has already crushed the reputation of Theory as we know it in the arts and humanities? What do you say to people who accuse you of promoting the Big Enemy of critical thinking?
YH: For me the main stake of Big Data, together with algorithms, is prediction. It is another form of the determination of time, which is probably not the same form of temporizing the past, the present, and the future that we can find in Bergson, Heidegger, Lyotard, Deleuze, etc. This means that we must discover in Big Data a new and powerful synthesis of time, and figure out how to deal with it. This new synthesis of time is what I call “tertiary protention,” which is intended to supplement Stiegler’s concept of tertiary retention. As we have discussed before, for Husserl there is primary and secondary retention, as well as primary and secondary protention (anticipation). In Stiegler’s theory, tertiary retention is the support for other forms of retention and protention; however, we must add that protention cannot be reduced to retention. This is very explicit in Husserl’s later writings on time-consciousness, e.g., the so-called Bernau manuscript (1917–18). Of course, there is ambiguity—for example, debt is an example of tertiary protention as well as tertiary retention, since it anticipates that which we will have to return, and it is recorded as traces. Tertiary protention is amplified due to the increasing ability of machines to predict and to anticipate. We might say that as long as we become part of Big Data, we are actually constantly in debt to certain unknowns.
We know the story of Edward Bernays and we know about the psychology of marketing, which since the twentieth century has been based on a mechanism geared toward the manipulation of psychopower. Now, however, the mechanism is not just concerned with psychopower; rather, personalization and prediction have become even more effective and direct. The predictions of Big Data give us an “average” experience, since Big Data is based on the mean. However, it is not average in the sense that everyone is the same; rather, Big Data shows variations around the mean, which give the impression that everyone is different. These variations are what Deleuze would call “the particular,” meaning that they can be reduced to a mean, to an average. They might also be described as the “differences” that sociologists Scott Lash and Celia Lury pointed out in their book Global Culture Industry. However, these differences are reducible.
Therefore, I would not say that Big Data is boring, but rather that it is truly suspicious, and we will have to transform this practice of Big Data. This is also related to your question of why the digital humanities haven’t risen up against this monstrosity. Many digital humanities projects are part of this paradigm. When you visualize the co-relations between hundreds of thousands of images, you are employing the same logic as the Big Data industry (albeit harmlessly) and you are exhibiting its aesthetics. This kind of digital humanities still has a place for now, but I don’t believe it can continue much longer, since we are reaching the end of a transitional stage. Data is by no means our “Big Enemy.” We should be aware of the history of data, which has been a subject in the humanities for a long time without being thematized. It is now time to enter a new stage by taking the question of data and the organization of data further. It seems to me that this has to be the task of the future “digital humanities.”
GL: You have said that “the digital is the capacity to process data.” Can we dig into that? This “dynamic” approach presumes that there is also a static view, of zeros and ones, in which the digital merely is. Is it an intolerable thought that data can just exist, without any context—data as such?
YH: There are not only two views, static and dynamic. There are different orders of magnitude, and each of these orders of magnitude can be seen as a reality in itself. The methodology of On the Existence of Digital Objectsincorporates such an understanding of orders of magnitude, which it is often used in epistemology. Therefore 0 and 1 is one order of magnitude, and data another. If we regard 0 and 1 as the only order of magnitude, we will be easily trapped in a metaphysical impasse. The philosopher Edward Fredkin has proposed what he calls a “digital ontology,” or “digital physics,” since he takes 0 and 1 as the foundation of being, like Thales’s water, Heraclitus’s fire, or Anaximander’s apeiron.
However, when we look at things from a phenomenological point of view, this digital metaphysics doesn’t do much except confirm Heidegger’s critique of technology: its essence is no longer technological but enframing (Gestell), and being is treated as a calculable standing reserve (Bestand). This is why I have proposed that we focus on the question of data as the main question of the digital. I take this insight also partly from Jacques Ellul. In fact, already in the 1970s, in his book Le système technicien
—a work that extended Simondon’s analysis of technical objects—Ellul observed that the totalization of systems was possible only because of the computer’s ability to process data.
You have asked, “Can data just exist, without any context”? I think the answer is yes, even without having to follow Quentin Meillassoux’s critique of correlationism. Firstly, we need to understand the history of the concept of data. Data is what is given, as the etymology of the Latin word datumsuggests. At the same time, it is sense data, which is also given—Husserl calls it das Gegebene. The French word for data, donnée, which is also the past participle of the verb “to give” (donner), retains this sense. We can say that in empiricist and transcendental philosophy, there are different ways of organizing data. For Hume, it is based on the rules of association (contiguity, resemblance, causality), and for Kant it is based on certain a prioristructures, including intuition and the understanding.
The use of the word “data” to designate computational information is only employed towards the end of the first half of the twentieth century. Essentially, this not only gives a new meaning to the term “data,” it also implies a necessity to rethink its organization. Hence the reason for this book. However, whether what is given is conceivable or not is another debate. When Heidegger talks about Being as es gibt, the word geben is emphasized as sending (schicken), as Geschenk, and what is given presents itself and hides at the same time, as Heraclitus says in his fragments. We might say that there is Datum an sich, like Kant’s Ding an sich, but it doesn’t necessarily mean that data is a black box or that it withdraws, as some speculative realists have said. For Heidegger, only through hiding is revealing possible. And even if we say that data belongs to the noumenal world, most Chinese philosophers would disagree with Kant that humans don’t have intellectual intuition and cannot access the noumenal. This is why I wanted to turn this dead-end question of “withdrawal” and Ding an sich into a question of relations.
GL: In the past, I learned to make a distinction between passive and active digital objects. There were executive files and static files such as documents or database entries. Does it make sense to make a distinction between programs and data? There is also a sociological dimension here: programs are written by geeks, whereas data is produced by clueless, ordinary users. These days, people talk about algorithms and bots. Both of them manipulate data in their own way.
YH: A long time ago, when we played games that came on floppy disks, it was necessary to use an .exe file to execute a .dat file. I guess this is what you mean by active and passive. This is still the case in some computational environments. The web, however, is a different environment, since it is supposed to be running all the time and is programmed in most cases with scripting languages. In general, in the past fifty years the mark-up languages have further developed and evolved—for example, from GML to SGML, from HTML1 to HTML5, from XHTML 1.0 to XHTML 2.0, and now web ontologies as well as formal ontologies. The use of mark-up languages like GML to format data started with IBM in the 1960s, and then in the 1980s there was a lot of work on knowledge representation (KR).
When we examine these histories, we see that the line between a data object and a program started to blur: not only do these objects carry constraints and functions, they also effectively allow communication between different platforms and applications. Programs and platforms can only communicate when the “ontologies” or “categorizations” are shared. They are becoming more and more “active” in the sense that you just spoke of.
GL: You write that the phenomenological tradition failed to comprehend technical and digital objects. At the same time, it is undisputed that Martin Heidegger is one of the most influential technology philosophers of the twentieth century. How do these two things go together?
YH: Let me be precise about this critique of phenomenology. I hold that the new definition of data seems to have problematized phenomenological investigations, which give an ambiguous role to technical objects in the construction of experiences. It is true that phenomenology has its own history dealing with technical objects in the larger sense of the term. For example, the early Husserl prioritizes expression (Ausdruck) over indication or sign (Anzeichen), since the latter doesn’t express anything—it is passive, like Hume’s association of ideas, while the former always demands an active sense explication. The late Husserl developed a different insight, where he addresses cultural objects, and the lifeworld (Lebenswelt) was primary in his investigation. Heidegger’s analysis of the ready-to-hand—which for me is actually a reversal of Husserl’s distinction between expression and indication—is very important to the understanding of technical objects, and that is why I offer it as a supplement to what Simondon calls the “concretization” of technical objects. I think that Simondon was aware of that, since he made Heidegger his ally in Part III of Du mode d’existence des objets techniques.
When I say that the phenomenological tradition is not sufficient to deal with digital objects, I mean first that the role of the technical object is ambiguous in these investigations, and therefore we must retrieve it through a rereading of Husserl and Heidegger—and here we must thank Jacques Derrida and Bernard Stiegler for their pioneering work (and we must also pay attention to the differences in their readings). Second, there is a reluctance to investigate the constitution of these objects. Husserl left what constitutes so called “pre-predicative experience” largely unexamined, surprisingly enough, considering that Husserl’s slogan was “back to things themselves.”​​
Phenomenology concerns the question of experience, which is how the subject constitutes itself through intentionality (whether via genesis or embodiment) and how objects are constituted as phenomena in the immanence of consciousness through intentional acts. To be more precise, there is a polar relation between the subject and the object, but what constitutes the object pole is rather limited, or maybe even only phenomenal. For example, phenomenology does not look into the schemes inside a technical object, and for this reason Simondon says that a phenomenological investigation of technical objects is dangerous. The investigation of digital objects is an attempt to rework the object pole and redefine its relation to the subject—that is to say, to experience. We must say that compared to Husserl, Heidegger paid much more attention to objects as well as to the constitution of objects. However, he did so in a different direction. Heidegger wanted to show that the constitution of the object is ontotheological, a tradition that started with Plato and Aristotle—though it is more complicated with the latter, since the early Heidegger’s lectures on Aristotle praised him for being closer to the Pre-Socratics than to Plato on the question of Being. A fiercer critique from Heidegger arrived later, for example in his four volumes on Nietzsche, in which Aristotle is described almost as a reactionary against Plato.
GL: From the very beginning data has had its own metadata. Files have names or a unique string of numbers. They go together. This is also what you say about digital objects: the “ontologies” are not separate from the actual data.
YH: Indeed, ontologies can be simply described as metadata schemes, which define and hence give meaning to data. Beware: the term “ontology” here is different from how it is randomly used in the humanities today. I describe this evolution of metadata schemes as a genesis of digital objects, and we can see that with the ontologies of the semantic web, descriptions of data are more refined, and the objectness of these entities becomes very clear. I remember already in 2010, during a conference on the semantic web, an engineer said that we were no longer dealing with mere data, but things, in the sense that data had become things. And if we pay attention to what this means, we see that it is not simply about how to do categorization—though categorization remains a crucial question and practice. It is also that categorization becomes productive. It produces objects in their own right, like Kant’s concepts, and these objects are both real and material. In this sense we can talk about the onto-genesis of digital objects.
GL: With Simondon, we could say that our efforts in media theory, electronic arts, tactical media, digital design, and net criticism can be described as a movement to reinscribe technics in culture. In most cases, however, they drift apart—not the least in philosophy itself. In today’s philosophy as (media) spectacle, we witness the authentic writer in the live act of deep thinking. Technology might spoil the party. Your genesis of digital objects might not be in high demand. Are you aware of that tension?
YH: I am not sure that what you have described can be called a movement to reinscribe technics in culture in Simondon’s sense, though I must admit that there is much excellent work that I appreciate a lot. According to Simondon, we need to overcome the opposition between culture and technics. This is because on the one hand, technology has been seen as a source of alienation, as what is responsible for the decline of culture; on the other hand, culture denigrates technics as something inferior in the social hierarchy. For example, robots are often seen as slaves—technical objects are only objects of consumption. For this reason Simondon, at the beginning of Du mode d’existence des objets techniques, says that his task is to show that “there is no such thing as a robot … a robot is no more a machine than a statue is a living being”; a robot “is merely a product of the imagination, of man’s fictive powers, a product of the art of illusion.” That is to say, we need a turn: it is not simply about studying technology, but rather turning technology into a support for culture. I’ve seen many researchers working on topics such as the sociality of Facebook or Twitter, but I’ve rarely seen any critical stance on this. As a result, the research becomes an added value to the industry—which also claims that it reinscribes technics in culture, but this is really just the culture industry. In philosophy, decades ago, we saw the tension between ontology and epistemology expressed in the legendary Davos philosophical debate between Heidegger and Ernst Cassirer in 1929. The former read Kant according to his fundamental ontology, while the latter rejected this reading and instead proposed an epistemological one. It is clear today that there is a fundamental tension between ontology and technics. In fact, this was already very clear in Heidegger’s fundamental ontology and in his analysis of modern technology, which for him was a consequence of Western ontotheology. Stiegler’s three-volume Technics and Time is important because it demonstrates this tension and suggests another framework for thinking this tension as not an opposition. However, there is still much work to be done to make this question more visible and to reflect on it in different domains.
GL: Relational technology plays an important role in your book. We could consider it the basis of all social media. Would it make sense to further develop a philosophy of the relational model?
YH: Yes, indeed, that is the principle question of my book. And for myself, the question of being is the question of relation. Over the years I have tried to work this out in a rereading of Heidegger, which I left out of the book so as not to obscure its object or message. We have seen that in recent years, some theorists have proposed certain relational models, but many of them do not specify what a relation is. I am not sure if one has to stroll through Whitehead’s Process and Reality in order to show that an app is relational. In my book, I try to answer the question: What is a relation? And what does it mean when we think of being in terms of relations, especially in the digital condition? The term “relation” has been used in semiosis and perception, but semiosis and perception don’t exhaust the question of relation.
In medieval philosophy, we have relationes secundum esse and relationes secundum dici, one according to being and the other according to speech. In my book I didn’t follow this vocabulary of medieval philosophy, since I wanted to move away from substance and theology, so I redescribed these relations as “existential relations” and “discursive relations.” I wanted to describe a dynamic model in which, firstly, both relations are in reciprocal relation, and secondly, technology can be seen as the process of the discovery (which is mostly the task of science) and materialization of discursive relations (this is the question of logos). As you can see in chapter three of the book, entitled “The Space of Networks,” I wanted to retrieve the concept of relation from Ancient philosophy, and then elaborate on the materialization of discursive relations; and in chapter four, “The Time of Technical Systems,” I reinscribe it in what I call a technical system, in which the discursive relations become inter-objective relations, and existential relations manifest themselves as temporalities. This is the general model that I propose for the analysis of technical systems, and I have used it in multiple practical projects. However, I must admit that it is impossible to exhaust the question of relation, and I will continue elaborating on it in future works.
GL: As an outsider to the main international standards organization for the World Wide Web, the W3C (World Wide Web Consortium), I have witnessed a move away from the semantic web towards a much more political aim of “re-decentralizing” the web, particularly in the post-Snowden period. Tim Berners-Lee was the original inventor of the web, back in 1991. His proposal for a new way to organize knowledge on the web, outlined in his 2001 article “The Semantic Web,” failed because of its inability to understand language (as Bernard Stiegler and others claimed). My interpretation would be that the naive multi-stakeholder approach got stuck in the monopolistic power politics of the stacks—Google, Facebook, Apple, and Microsoft—which demonstrated that they were uninterested in the formalistic, scientific rearrangement of protocols. In the end, the scientists were pushed aside.
YH: I was very interested in the semantic web, both its logical questions and philosophical implications. In 2010, along with Harry Halpin and Alexandre Monnin, we launched the program “Philosophy of the Web” in Paris, which consisted of various events. I still think the semantic web is a very important project in the history of the web. The semantic web was intended to be a “world-building” project, and this is the reason Tim Berners-Lee called for “philosophical engineers,” who would not only reflect on the world but build the world—an echo of Marx’s thesis on Feuerbach. The semantic web aims for a world of automation. However, a world is more than automation; it also has politics, which the semantic web doesn’t take into account. I don’t think this is because the semantic web doesn’t understand language—and we have to admit that machines don’t deal with language in the way we do. This is why I suggest that we surrender the opposition between syntax and semantics and instead take up the concept of relation.
Brian Cantwell Smith, in his early and very important work On the Origin of Objects, has a very nice argument against the claim that machines only have syntax and no semantics, since such a distinction is far too anthropocentric. Contrary to what you have said, I am rather sure that Google, Facebook, Apple, and Microsoft are all interested in “the formalistic, scientific rearrangement of protocols”; however, they all want their own protocols, and so they are reluctant to all use the same standards. We have to recognize that there is an institutional politics between the W3C and its business members. I think someone who looked more deeply into the history of the W3C would have better insight on this. It is true that since the Snowdon affair, the W3C has launched the Magna Carta project and the campaign “Web We Want.” However, since its launch it doesn’t appear to me that there has been much progress.
The other reason for the “failure” that we have described—and Stiegler has been claiming this for years—is that the semantic web did not allow for a “social web,” since its ultimate aim was the automation and standardization of data schemes. This is a different issue than the “cyber-libertarian” project of Julian Assange. Rather, it is a question of social organization and the organization of the social. To address this question of automation, in my book I attempted to compare Husserl’s intentional logic with extensional logic in order to show that we should reintroduce the question of experience into formal logic. This stands out as a rather strange chapter in the book, since it proposes a reading of Husserl that is closer to Deleuze and Simondon. This requires a long detour through Frege, Hilbert, Kripke, and Putnam. In 2012, I worked with Stiegler and Harry Halpin to reconceptualize the concept of the social by departing from Simondon’s notion of collective individuation in order to develop an alternative to Facebook. Just as Uber is the biggest taxi company without taxis, social networks are the biggest communities without the social. The semantic web only wants to provide an industrial standard so that these industrial players will use it to facilitate the development of the web, to avoid “walled gardens,” as some have said. But advocates of the semantic web have nothing to say about the industry itself. This is the stake of the semantic web, and not its failure to understand language.
GL: Let’s end with your upcoming book on the status of technology in China. Can we see this as a follow-up or logical extension of On the Existence of Digital Objects? Has your decade in Europe made it easier to reflect on China? What do you make of people who travel to Shenzhen to do ethnography there? Can philosophy be the king or queen of the sciences and in this way beat the social sciences?
YH: Indeed, the new book is intended to be a second work on the concept of relation that we discussed earlier. In On the Existence of Digital Objects, I deal with formal relations and objects. In The Question Concerning Technology in China: An Essay in Cosmotechnics (Urbanomic 2016), I deal with the relation between the cosmos and the moral. This book on China is an attempt to elucidate the differences between the way the concept of technics is understood in Chinese philosophy and the way it is understood in Ancient European philosophy. And as the title suggests, the book is an attempt to recontextualize and problematize Heidegger’s famous essay “Die Frage nach der Technik,” in order to revive the concept of a technics of world history, which I call “cosmotechnics.” Picking up what François Jullien says, we can know ourselves by knowing others. His work on Chinese thought allows him to better understand European thought. I profited from years of living and studying in Britain, France, and Germany, reflecting on different systems of thought. A few years ago you joked that I was actually doing ethnography in Europe. With this book, I want to show that there has been a different concept of technics in China. It is neither the Greek technē, nor “technology” in the sense that emerged in European modernity. This difference is not obvious among researchers in China, and it has never been clearly articulated; indeed, this was very embarrassing! I once read an article from a well-known Chinese philosopher of technology who, when addressing the Chinese public, claimed that Prometheus was the origin of all technics (including Chinese technics). That is a complete disorientation, in the double sense of the word. Maybe the Greeks and the Chinese all come from Prometheus, but this is not easy to prove …
I am probably not the best person to comment on the debate between philosophy and the social sciences. I wouldn’t say that there is a king or queen of disciplines. However, we have to acknowledge that in philosophy there is a particular form of questioning and a strong attention to histories of thought and to the precision of concepts. This way of questioning allows us to problematize a lot of dubious definitions that are often taken for granted. I am also interested in the social sciences, and my first degree was in computer engineering with a focus on AI, and I continue to work on practical projects. Any insistence on the superiority of a discipline is in most cases only self-indulgence. Early this year in Berlin I spent thirty minutes listening to Alain Badiou and Jean-Luc Nancy debate the question of whether Marx was a philosopher. I wish I could get those thirty minutes back. I don’t see what more we could get out of Marx if we renounced him as a philosopher. The rigor of a work is not solely determined by institutions or tradition. It depends on historical insights, consistent interrogations, and creativity. There is bad social science just as there is bad philosophy, not to mention bad scientific research.
Apropos of Badiou, recently he criticized Pokémon GO as “the corruption of corruption” and claimed that “the battle against images is a Platonic battle.” It is astonishing that this came out of the mouth of a Maoist, since every French Maoist knows by heart the saying “No investigation, no right to speak.” However, we must also turn the question around: How deeply must one engage with Pokémon GO in order to speak about Pokémon GO? Or more generally, how deeply must one understand technology in order to talk about technology? We easily fall into two extreme orders or two problematic philosophical attitudes: one simply renounces modern technology, since it is intrinsically bad; and the other dogmatically endorses it in order to endow it with a certain “ontological dignity.” We should get out of this Unmündigkeit, as Kant would call it, and overcome these obstinate oppositions. What is denounced may always appear in other forms in those who denounce it.
I hope that my book on China and technics can at least remind researchers who are, in your words, “doing ethnography in Shenzhen,” that in China there is a history of technics and a history of modernization. Some researchers take globalization as a given fact so they can simply study the differences between “technical facts”—in André Leroi-Gourhan’s sense, meaning the specificities of the tools and the different gestures of their users—without looking into the history of technics and modernization in China, into their “form of life,” as if China is no different from an African country, or as if the differences that do exist are only superficial. Ethnographers know very well that one must problematize globalization and modernization. We may want to remind ourselves that after having witnessed the disintegration of nonmodern cultures, Claude Lévi-Strauss addressed his fellow anthropologists in Tristes Tropiques by saying that anthropology should be renamed “entropology.” However, some quasi-critical ethnographic works only nurture such modernization. While we don’t expect everyone to be Joseph Needham and we don’t want to operate on a simple opposition between the global and the local, but do have to recognize “ontological diversities,” as has been proposed by Philippe Descola, Eduardo Viveiros de Castro, Bruno Latour, and others who are part of the so-called “ontological turn” in anthropology. This is why I believe that, besides the proposal by these anthropologists to recognize multiple natures, we must first of all recognize the diversity of cosmotechnics, without which there is no discourse of nature—diversity not only in the sense of different “technical facts” or “technical systems” (as Leroi-Gourhan and Bertrand Gille have put it) but also in the sense of different ontologies and cosmologies. And once this multiplicity is affirmed, how are we going to imagine the development of technologies and theories in the Anthropocene? This will be the next battle for all of us.
×
Yuk Hui studied Computer Engineering, Cultural Theory, and Philosophy at the University of Hong Kong and Goldsmiths College in London, with a focus on philosophy of technology. He is currently research associate at the ICAM of Leuphana University Lüneburg. Yuk Hui is co-editor of 30 Years after Les Immatériaux: Art, Science and Theory (2015), and author of On the Existence of Digital Objects (prefaced by Bernard Stiegler, University of Minnesota Press, March 2016), The Question Concerning Technology in China. An Essay in Cosmotechnics (Urbanomic, December 2016).