线上言论规则的奇异曲折之路 （ technology tech law)

公共领域中有很多关于社会线上言论应用的程度和法律条律种类的辩论。过去数年中对话变得更加紧张和急迫，对网络言论施加限制的努力从一开始就是倍受争议的。如今，网络礼仪行动的片段230正经历着再审查。

在言论规则中选取下意识的立场是容易的，但他们在历史理解中通常并不成为一个圆满或持续的结果。事实上，正是由于仓促的判断和对互联网绝对新颖性的模糊理解，才使我们首先来到了因此我想概括简明地陈述互联网规则历史：抛砖引玉。我对论述了这一处理的 Hal Abelson, Ken Ledeen and Harry Lewis的书籍 Blown to Bits有很多研究。

广袤的野生网络

让我们追根溯源，但不会在此费很多篇幅。互联网作为一个由高级企划研究所（或ARPA）运行的军事研究项目，开始于19世纪60年代，自那时被更名为 “ DARPA” （D意为防御）。它被设想为一个中断阻力的对话媒介，这样策略性置放电子通信的节点不会阻止被发送的信息。

确切地说，它是电话线等传统电信的替代品，如果苏联将合适的城市夷为平地，电话线将是失败的。

互联网杰出地完成了它的目标，并且仍在继续着它的使命。只要源头和目的地的路径存在，它就能毫不费力地在飞行中改变除中心架构外的数据包，将它们送往去处，

为了检测这个，ARPA与国内最有声望的大学和调查公司合作，并将他们紧密地起来。互联网上曾一度只有这些组织的人，如下图所示。

政府雇佣的计算机科学家们在家中连上了网的第一批人中。很多第一批次的黑客，和“网络空间”的前辈们，都是这些专家的孩子，用他们父母的设备漫游早期网络的公告栏。

电脑真正在消费者中扩散是在19世纪90年代。电子邮件开始成为美国成年人日常生活的一部分。然而，到消费者和关键的立法者第一次遇到需要规范的线上不良行为时，粗鲁的全盘自由传统已经生根发芽。

公告栏服务习惯在没有干预的情况下运行，并且网络服务供应商 （ISPs）乐于提供字节并将剩余部分留给其他人使用。

长期用户不愿放弃他们的消费品味自由，与消费者和政客们强烈抗议的不协调，被少数人对网络的滥用所震惊，引起了针对网络规则的鞭挞。

镇子里有一个治安官

互联网上确实有两种影响着演讲的内容：诽谤性内容和淫秽内容，尤其是那些可能伤害儿童的内容。在模拟世界中，许多不同立场的人必须共同努力以促进言论表达，但他们对令人反感的言论承担不同程度的法律责任。

作者总是承担最大的责任，因为演讲是他们自己的。出版商也负有责任，因为他们对作者的话语拥有编辑控制权，这意味着他们知道发表文章的作者在说什么，并在上面签字。

发行人通常不被判罪，因为他们通常不知道，也不期望知道他们所发行的内容。想想送报纸的孩子：阅读报纸并确保其中没有虚假或淫秽的内容不是他们的工作。

内容生产链中各方的分工看起来足够合理和直观，但是立法者，法官和互联网用户发现，将他们应用于互联网绝非易事。

在试图将互联网内容设备弯曲成类似于模拟设备的模型时，立法者通常只能监管少数几个方。

他们可以监管居住在美国管辖范围内的作者。另外，他们可以规范作者的ISP，但前提是它必须在美国管辖范围内运营。最后，立法者还可以选择基于消费者在美国的假设来规范消费者的ISP和消费者自己（就像美国人在触及美国法律时，因假定的居住地在美国，而从美国的反诽谤和反淫秽法令中受益一样）

1991年，两个在线公告板服务商之间发生法律纠纷，这标志着美国法院第一次肯定地向在线播放器添加了分类（作者，发行者或发行者）。当时，CompuServe公司维护了一个名为Rumorville的谣言论坛，该论坛发布了第三方提供的内容。一个关键的细节是，CompuServe不审核收到的任何材料-它只发布生产者提供的合同内的任何内容。

另一电子公告牌运营商Cubby支持了Rumorville的竞争对手Skuttlebut。不久之后，在Rumorville上谣言四起，声称Skuttlebut是伪造的，并且由于Cubby认为事实是CompuServe散布虚假信息以打击它，Cubby起诉了CompuServe涉嫌诽谤。

作者在“ Brown to Bits”中这样对案件进行描述：“为了更好地进行类比，法院将CompuServe描述为“电子营利性图书馆”。作为发行人或图书馆，CompuServe独立于[内容创建者]，对[创建者]提供的诽谤言论不承担任何责任。CompuServe很快解决了Cubby起诉CompuServe的案子

换句话说，尘埃落定后，在线平台就被视为发行商，这意味着他们与其用户或提供商提供的任何令人反感的内容不再有所关联。

这正是下一个具有里程碑意义的法院案件震惊在线平台的原因。它的开端与Cubby起诉CompuServe的案件几乎相同，公告板遭到诽谤诉讼的打击。 1994年，Prodigy旗下以财务为重点的董事会Money Talk上的一位匿名用户指控Stratton Oakmont公司为“重大犯罪欺诈”。斯特拉顿·奥克蒙特（Stratton Oakmont）起诉Prodigy诽谤罪，导致Stratton Oakmont诉Prodigy。

该案例带有一个转折点：Prodigy渴望在董事会上营造一种家庭般友善的氛围，并公开宣称，它调整了平台以清除色情内容。法院认为该细节令人信服，并裁定原告胜诉。

Abelson等人写道：“通过行使编辑控制以支持其家庭友善的形象，Prodigy成为出版商，并承担责任和风险。”

对于法院而言，事实检查超出了Prodigy的控制范围并不重要。如果平台完全主持，它将担当编辑角色，这将使其对托管的所有内容承担责任。因此，该决定不鼓励公告牌服务部门承担任何编辑职责，以免他们发现自己与令人反感的内容挂钩。

完美的激流

如果我说，一项在道德上令人怀疑的科学研究，耸人听闻的新闻和过分热心的参议员导致了有史以来最有影响力的互联网言论法，你会相信吗？

这听起来很奇怪，但的确是真实发生的事情。

1995年7月3日，《时代》杂志刊登了一个令人震惊的封面曝光活动“ CYBERPORN”，并立即在华盛顿引发了道德恐慌。后来发现，故事背后的研究人员使用了不道德的方法。

例如，他们补偿参与者提供的材料加强了作者的偏见（互联网上充斥着色情），并伪装成同行评审期刊的发表了他们的作品，但只比一篇学生论文得到的报酬略多。

那些试图阻止亵渎网络内容浪潮的国会议员发现自己陷入困境，因为在Stratton Oakmont诉Prodigy案之后，没有人希望触及内容审核。最后，1996年《通信道德法》（CDA）获得几乎全票通过。

该法律有两个关键组成部分-“展示规定”和“善意撒玛利亚规定”。

“展示规定”将打击目标置于可能会出现在孩子面前的屏幕上的“淫秽不雅”内容上。此外，“展示规定”提出：“任何交互式计算机服务皆展示以18岁以下群体可接受方式，在上下文中出现的任何评论，请求，建议，提议，图像或其他通信以当代社区的标准衡量，以冒犯性的方式描述或描述性或排泄活动或器官则视为非法”。此外，该法律还包括一项良好的撒玛利亚条款，该条款构成了第230节的核心（通常援引“ 230节”作为参考）。它的语言完成了两项重要的壮举。

首先，它保护任何“交互式计算机服务”免于因努力消除“淫秽，淫荡，淫荡，肮脏，过度暴力，骚扰或其他令人反感”内容而造成的后果。

其次，它在物质世界的类似情况中将这些服务归类为“分发者”：“任何交互式计算机服务的提供者或用户，均不应被视为另一信息内容提供者提供的任何信息的发布者或说话者。”

ACLU在ACLU诉Reno案中对CDA提出质疑，理由是CDA过度限制了第一修正案的合法演讲，并取消了关于展示的规定，因为这违反宪法。

正如斯图尔特·达尔泽尔法官在多数意见中所指出的那样，“要求所有可能看到成年人合法可见材料的人都通过互联网进行年龄验证，是令人无法接受的。”

辩方案的症结在于应按照监管电视的方式对监管互联网，但法院认为这种比较是不恰当的。

“互联网被公认为是一场永无止境的全球性对话。政府可能不会通过CDA中断该对话。作为迄今为止最具参与性的群众发言平台，互联网应受到最高保护，以免受政府的操控。”达尔泽尔写道。

善良的撒马利亚人，可疑的好结果

值得注意的是，第230节的“善良撒玛利亚言语言“在ACLU判决中保持不变。尽管这是现在公众讨论的规章制度，该案并没有在第一次就引发争议。确实，问题立刻就出现了。

1997年，马特·德拉吉（Matt Drudge）在美国在线（AOL）上发布了关于时任克林顿总统助手的西德尼·布卢曼塔尔（Sidney Blumenthal）的诽谤性言论，对法律进行了一次考验。尽管AOL对其发布的Drudge材料具有编辑影响力，但法院裁定该公司不是发行商，因此对诽谤不承担任何责任。该意见引用了CDA的“善良撒玛利亚语言”。

1998年，Jane和John Doe（在本例中为一个母亲和她的儿子）起诉AOL，因为它允许用户出售未成年人John制造的色情材料。在其用户协议中，AOL保留终止任何存在滥用行为的用户服务的权利。在这种情况下，仍引用了《善良撒玛利亚人》条款免除美国在线的责任。

Abelson等总结问题道：“国会在发表错误的表率之后，不提供任何正确的类比，这使混乱成为可能。”

赛伯兰法则

完成此历史概述后，我们或多或少地了解了当前的技术时代。

第230条仍然是少数几个鼓励在线平台之间进行内容审核的力量之一。但致命的缺陷是，只要他们能够令人信服地证明自己的行为是真诚的，他们行为的后果就不会受到法律制裁。

当然，已经确定的是，如果线上服务收到联邦犯罪或知识产权盗窃通知，并且不采取任何行动，那它们将失去责任保护，但第230条几乎是绝对的。

因此，在线平台具有创建和执行他们选择的社区标准的广泛自由。如果标准的执行过多，不足或不合时宜，则平台的运行人员可以躲在诚实信用辩护的后面，天真地声称没有人是完美的。

言论自由的拥护者认为，这可能是正确的，不应盲目裁决，武断地决定谁可以在平台上，以何种方式“发言”。

正如我想说的那样，另一个促成因素是“网络上没有人行道”。几乎整个Web（特别是与Internet不同的Web）都是私有财产。 《第一修正案》禁止政府审查美国人的言论。

例如，政府无法告诉您在您站在人行道时可以说什么和不能说什么（除了一些公共安全方面的例外），因为人行道是公共财产。但是，第一修正案不适用于大多数网络平台所使用的私人实体。如果你注册了社交网络，则默认你同意互联网的规则，包括那些禁止某些言论的规则。

就像80年代和90年代的网络由先驱者建立起来只是为了让立法者气喘吁吁，开拓者一直在努力推动着，使公务员再度陷入尘埃。90年代后期认为足够强大的措施在互联网上应用，在新的压力和更复杂的使用模式下开始呻吟。

作为一个社会的成员，跟上它的步伐是困难的，我们只能尽力而为，这需要为我们来到这里而感到自豪。另一种选择是根据一时冲动做出决策，这些决策不太可能经受时间的考验，而且很可能在过程中造成严重破坏。

The Strange, Meandering Journey to Online Speech Regulation

There has been a lot of debate in the public sphere around the degree and kind of legal regulation a society should apply to online speech. While the dialogue has become more intense and urgent in the last few years, the effort to impose limits on Internet speech has been contentious from the start. At the present juncture, Section 230 of the Communications Decency Act is undergoing reconsideration.

It's easy to take knee-jerk stances on Internet speech regulation, but they generally do not achieve as satisfactory or sustainable end results as stances that are grounded in an appreciation of history. In fact, it is precisely hasty judgment and foggy understanding of the Internet's sheer novelty that got us to this fractious juncture in the first place.

That's why I want to present a brief overview of Internet regulatory history: to do my modest part to set the conditions for more enlightened outcomes. I owe much of the research that informs this treatment to a book called Blown to Bits, by Hal Abelson, Ken Ledeen and Harry Lewis.

If you are interested in getting a fuller, but still digestible, understanding of how radically new and unprecedented the Internet is, it is worth checking out, which you can do for free (it's licensed under Creative Commons).

The Wild Wide Web

Let's start at the beginning, but we won't spend too much time there. The Internet began in the 1960s as a military research project run by the Advanced Research Projects Agency, or ARPA, which since has been renamed "DARPA" (the "D" standing for "Defense"). It was devised as an outage-resistant communications medium, so that the downing of strategically placed telecommunications nodes would not prevent messages from being sent.

To be precise, it was meant to be an alternative to traditional telecommunications, like telephone lines, which would fail if the Soviet Union leveled the right city.

The Internet accomplished its goal brilliantly, and it still does what it was built to do. It effortlessly reroutes data packets on the fly, without a centralized architecture, to get them where they're going as long as any path between the source and destination exists.

To test this, ARPA partnered with the country's most prestigious universities and research firms and linked them all together. For a while, the only people on the Internet were the researchers at those institutions, as this 1970 map of the Internet shows.

By the 1980s, the Internet opened to the public, but it was so arcane and inaccessible that only a small cadre of private and public sector players -- and, as it turned out, their family members -- had any contact with it. Home computers ran in the thousands of dollars, making them impractical to most but obtainable for some. Large corporations had started using the Internet as part of their operations, so some of their employees followed suit at home to get in more practice.

Government-employed computer scientists were among the first to have Internet-connected devices in the home. A lot of the first wave of hackers, as well as progenitors of "cyberspace" culture overall, were the children of those professionals, who used their parents' devices to rove around the bulletin boards of the early Web.

Computers really proliferated among consumers in the 1990s. Email started to be a regular part of the lives of American adults. However, by the time consumers -- and crucially legislators -- first encountered abusive online behaviors that merited regulation, a robust tradition of total freedom already had taken root.

Bulletin board services were accustomed to operating without interference, and Internet service providers (ISPs) were content to deliver the bytes and leave the rest to someone else.

It was the dissonance between the reluctance of longtime users to give up their taste of freedom and the outcry from consumers and politicians appalled by the abuses of a few that begot the whiplash in Internet regulation.

There's a Sheriff in Town

There were really two types of content that shaped speech on the Internet: defamatory content and obscene content, especially any that could harm children. In the analog world, many different parties must work together to facilitate the expression of speech, but they bear different degrees of legal responsibility for objectionable speech.

Authors always bear the greatest responsibility, since the speech is their own. Publishers are also responsible, because they wield editorial control over the author's words, meaning they know what the authors they publish are saying and, by extension, signed off on it.

Distributors generally aren't held culpable, because they usually don't know, and aren't expected to know, the content they are distributing. Think of newspaper delivery kids: It's not their job to read the newspaper and make sure it doesn't contain any falsehoods or obscenities.

These categorizations of parties in the content production chain seem reasonable and intuitive enough, but what lawmakers, judges and Internet users found was that applying them to the Internet was no simple matter.

In trying to bend the Internet content apparatus into a shape resembling the analog one, lawmakers generally can regulate only a few parties.

They can regulate authors who reside within U.S. jurisdiction. Alternatively, they can regulate the author's ISP, but also only if it operates within U.S. jurisdiction. Finally, lawmakers also have the option of regulating the consumer's ISP and consumers themselves, based on the assumption that the consumers are in the U.S. (just as Americans benefit from U.S. anti-defamation and anti-obscenity statutes due to their assumed physical residence within the reach of U.S. law).

A legal scuffle between two online bulletin board services in 1991 marked the first time that U.S. courts affirmatively affixed a classification -- author, publisher or distributor -- to an online player. Back then, the company CompuServe maintained a rumor forum, Rumorville, which posted content provided by third parties. The key detail is that CompuServe did not review any of the material it received -- it merely posted whatever its contracted content producers provided.

Another bulletin board operator, Cubby, propped up Skuttlebut, a competitor to Rumorville. Shortly afterward, a rumor cropped up on Rumorville alleging that Skuttlebut was phony, and because Cubby saw this as CompuServe spreading falsehoods to edge it out, Cubby sued CompuServe for defamation.

In Blown to Bits, the authors characterize the case this way: "Grasping for a better analogy, the court described CompuServe as 'an electronic for-profit library.' Distributor or library, CompuServe was independent of [its content creator] and couldn't be held responsible for libelous statements in what [the creator] provided. The case of Cubby v. CompuServe was settled decisively in CompuServe's favor."

In other words, when the dust settled, online platforms were deemed to be distributors, meaning they were off the hook for any objectionable content their users or providers transmitted via their platform.

That's why the next landmark court case took online platforms completely by surprise. It started out in much the same way as Cubby v. CompuServe, with a bulletin board getting hit with a libel suit. In 1994, an anonymous user on Money Talk, a finance-focused board owned by Prodigy, accused the firm Stratton Oakmont of "major criminal fraud." Stratton Oakmont sued Prodigy for libel, begetting Stratton Oakmont v. Prodigy.

That case came with a twist: Eager to engender a family-friendly atmosphere on its boards, Prodigy openly advertised that it moderated its platforms to scrub them of obscene content. The court found that detail compelling, and it ruled in favor of the plaintiff.

"By exercising editorial control in support of its family-friendly image, said the court, Prodigy became a publisher, with the attendant responsibilities and risks," wrote Abelson et al.

To the court, it did not matter that fact-checking went beyond the scope of Prodigy's intentions through its moderating. If a platform moderated at all, it took on an editorial role, which would make it liable for anything and everything it hosted. Thus, the decision discouraged bulletin board services from taking on any editorial duties, lest they find themselves on the hook for objectionable content.

The Perfect Torrent

Would you believe me if I told you that an ethically suspect scientific study, sensational journalism, and overzealous senators led to the most influential Internet speech law ever passed?

Strange as it sounds, that's exactly what happened.

A shocking cover exposé, "CYBERPORN," was published in Time magazine on July 3, 1995, and it immediately set off moral panic in Washington. It later emerged that the researchers behind the study at the heart of the story used unethical methodologies.

For example, they compensated participants for providing material that reinforced the author's confirmation bias (that the Internet was rife with pornography), and published their work in what masqueraded as a peer-reviewed journal but was little more than a student paper.

Those in Congress seeking to stem the tide of profane Web content found themselves in a jam, because after Stratton Oakmont v. Prodigy, no one wanted to touch content moderation. The result was the near-unanimous passage of the Communications Decency Act, or CDA, in 1996.

The law had two key components -- the "display provisions" and the "Good Samaritan provision."

The display provisions took aggressive aim at "obscene and indecent" content that could end up on a screen in front of a child. Among other things, the display provisions made it illegal for "any interactive computer service to display in a manner available to a person under 18 years of age, any comment, request, suggestion, proposal, image, or other communication that, in context, depicts or describes, in terms patently offensive as measured by contemporary community standards, sexual or excretory activities or organs."

Additionally, the law included a Good Samaritan provision, which makes up the heart of Section 230 (and which "Section 230" generally is invoked to reference). Its language accomplishes two significant feats.

First, it shields any "interactive computer service" from liability for the consequences of making good faith efforts to remove "obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable" content.

Second, it classifies these services as "distributors" in the meatspace analogue: "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."

The ACLU challenged the CDA in ACLU v. Reno on the grounds that it unduly limited legitimate First Amendment speech, and got the display provisions struck down as unconstitutional.

As Judge Stewart Dalzell stated in the majority opinion, "It would chill discourse unacceptably to demand age verification over the Internet from every person who might see material that any adult has a legal right to see."

The crux of the defense's case was that the Internet should be regulated the way TV is, but the court rejected the comparison as inapt.

"The Internet may fairly be regarded as a never-ending worldwide conversation. The Government may not, through the CDA, interrupt that conversation. As the most participatory form of mass speech yet developed, the Internet deserves the highest protection from governmental intrusion," Dalzell wrote.

Good Samaritans, Questionably Good Outcomes

Notably, the Good Samaritan language of Section 230 was left intact in the ACLU decision. Although this is the regulation public discourse now swirls around, that case did not mark the first time it sparked controversy. Indeed, problems emerged immediately.

One test of the law played out in 1997 when Matt Drudge posted allegedly defamatory statements about Sidney Blumenthal, an aide to President Bill Clinton at the time, on AOL. Though AOL had editorial influence over the Drudge material it posted, the court ruled that the company was not a publisher, and therefore was not liable for libel. The opinion cites the CDA's Good Samaritan language.

In 1998, Jane and John Doe (in this case, a mother and her son) sued AOL because it allowed a user to sell pornograhic material made of John when he was a minor. In its user agreement, AOL reserved the right to terminate service for any user who engaged in abusive behavior. The Good Samaritan provision also was cited in that case to absolve AOL of responsibility.

Abelson et al. sum up the problem: "Congress had made the muddle possible by saying nothing about the right analogy after saying that publishing was the wrong one."

The Law of the Cyberland

With this historical overview complete, we are more or less caught up to the current technological epoch.

Section 230 remains one of the few forces incentivizing content moderation among online platforms. The fatal flaw here is that so long as they can argue convincingly that their actions were in good faith, they are immune to legal consequences.

Granted, it has been established that online services lose their liability protection if they are notified of the commission of federal crimes or intellectual property theft and take no action, but Section 230 is nearly absolute otherwise.

As a result, online platforms have wide latitude to create and enforce whatever community standards they choose. If the speech standards enforcement is excessive, deficient or lopsided, a platform's operator can hide behind the good faith defense, innocently claiming that nobody's perfect.

True as that may be, free speech advocates contend that should not serve as a blank check to decide arbitrarily who can "speak" on a platform, and on what terms.

The other contributing factor is, as I like to say, "there are no sidewalks on the Web." Nearly the whole of the Web -- specifically the Web, as distinct from the Internet -- is private property. The First Amendment restrains the government from censoring Americans' speech.

Because sidewalks, for example, are public property, the government can't tell you what you can and can't say while you're standing on one (with a few exceptions for public safety). However, the First Amendment does not apply to private entities, which is what most Web platforms are. If you register to a social network, you consent to its rules, including those that prohibit certain kinds of speech.

Just as the Web of the 80s and 90s was settled by pioneers, only for lawmakers to catch up gasping for breath, the trailblazers have kept pushing on to leave civil servants in the dust once again. Measures that seemed robust enough in the late 90s are beginning to groan under the weight of newer and more sophisticated usage patterns on the Internet.

Daunting as it is to keep up, as members of a society we must do our best, which requires an appreciation of how we got here. The alternative is to make decisions on the spur of the moment, which are unlikely to withstand the test of time, and too likely to wreak havoc along the way.

Author: Jonathan Terrasi

https://www.technewsworld.com/story/86711.html