DRAFT : November 5, 2001

Against ‘Against Cyberanarchy’

David G. Post[*]

In 1998, the University of Chicago Law Review published Professor Jack Goldsmith’s “Against Cyberanarchy.”[1] In that article, Goldsmith vigorously challenged a number of claims made by those he called “regulation skeptics”: claims that “the application of geographically based conceptions of legal regulation and choice of law to a-geographical cyberspace activity either makes no sense or leads to hopeless confusion,”[2] that “because cyberspace transactions occur ‘simultaneously and equally’ in all national jurisdictions, regulation of the flow of this information by any particular national jurisdiction illegitimately produces significant negative spillover effects in other jurisdictions,”[3] and that “cyberspace participants are much better positioned than national regulators to design comprehensive legal rules that would both internalize the costs of cyberspace activity and give proper notice to cyberspace participants.”[4]

In the years since its publication, “Against Cyberanarchy” has had a major impact on the cyberspace debate. It has become a staple of cyberspace law courses[5]; it has been cited dozens of times in the law review literature,[6] and influential commentators have used words like “trenchant,”[7] a “withering” critique of the regulation skeptics’ claims,[8] a “decisive response to the widespread view that cyberspace is not regulable,”[9] to describe it.

Inasmuch as some of my own work was in the center of Goldsmith’s bulls-eye, I have read, and re-read, his article with some care. My friends and colleagues eagerly awaited – at least, they told me that they eagerly awaited – ‘my side’ of the story. I began, any number of times, to write the definitive response.

But it has been about 4 years now, and my silence has become deafening. Nothing I tried to set down on paper quite ‘worked.’ I found myself deeper and deeper in a very dark forest – the mysterious world of international conflicts of law and international jurisdiction. This, of course, is Goldsmith’s home territory, and I felt like the Hessians must have felt when they tried fighting on American soil, finding snipers and booby traps lurking around each corner. I couldn’t get Goldsmith’s arguments squarely in my sights, and, whenever I felt I had done so to my satisfaction, I found I didn’t have much to say about them. I guess Goldsmith’s critique really had been withering; I felt as though I had withered.

But I have, finally, come to the source of the problem. The argument that Goldsmith and I are having is not, I now realize, an ‘argument’ at all, at least not in the ordinary sense of that term. We may not disagree, really, about anything having to do with international conflict of laws, or international jurisdiction, or the Hague Convention on the Enforcement of Foreign Judgments, or the principles set forth in the Restatement (Third) of the Foreign Relations Law of the United States, or anything else that we might have occasion to discuss in our classrooms. It’s much simpler than that: we start from completely different, and probably irreconcilable, premises about the nature of the world.

The Unexceptionalists

“Transactions in cyberspace involve real people in one territorial jurisdiction either (i) transacting with real people in other territorial jurisdictions or (ii) engaging in activity in one jurisdiction that causes real-world effects in another territorial jurisdiction. To this extent, activity in cyberspace is functionally identical to transnational activity mediated by other means, such as mail or telephone or smoke signal.”[10]

Border-crossing transactions have always presented the international legal system with difficult and challenging questions of a “jurisdictional” nature (broadly defined): Whose law applies to such transactions? Which sovereign(s) have ‘jurisdiction to prescribe’ law for transactions that “originate” in one country and “terminate” elsewhere? When and to what extent is “extraterritorial regulation” permissible? Goldsmith’s position – what I will call “Unexceptionalism”[11] – is straightforward: that while activity in cyberspace presents many difficult and complicated problems of this type, these problems are no more difficult or complicated because the transactions take place “in cyberspace.”[12] Cyberspace or real-space – to the Unexceptionalists, it doesn’t really matter; there’s nothing truly “new” here, nothing about cyberspace that raises any difficult or troublesome issues that we haven’t seen before.

Take, for instance, Goldsmith’s description of a paradigmatic cyberspace jurisdictional problem:

“[C]onsider the predicament of one of the scores of companies that offer, sell, and deliver products on the World Wide Web. Assume that the web page of a fictional Seattle-based company, Digitalbook.com, offers digital books for sale and delivery over the Web. One book it offers for sale is Lady Chatterley's Lover. This offer extends to, and can be accepted by, computer users in every country with access to the Web. Assume that in Singapore the sale and distribution of pornography is criminal, and that Singapore deems Lady Chatterley's Lover to be pornographic. Assume further that Digitalbook.com's terms of sale contain a term that violates English consumer protection laws, and that the publication of Digitalbook.com's Lady Chatterley's Lover in England would infringe upon the rights of the novel's English copyright owner. Digitalbook.com sells and sends copies of Lady Chatterley's Lover to two people whose addresses (say, and ) do not reveal their physical location but who, unbeknownst to Digitalbook.com, live and receive the book in Singapore and London, respectively.”[13]

This scenario, Goldsmith and I agree, certainly raises some complicated problems. Does English law, or Singaporean law, or both, apply to Digitalbook.com’s conduct? Would application of either of these bodies of law constitute “impermissible extraterritorial regulation of a U.S. corporation”[14]? If Digitalbook.com “cannot limit its cyberspace information flows by geography,”[15] would application of English, or Singaporean, law cause Digitalbooks.com “to withdraw Lady Chatterley’s Lover”[16] (or, at the very least, to “raise its price”[17]), thereby “adversely affecting the purchasing opportunities of parties in other countries”[18]? And if so, are these “negative spillover effects”[19] of national regulation “illegitimate [and] unfair”[20] – especially given that “Digitalbook.com had no way of knowing that it sold and delivered a book to consumers in these countries”[21]?

However complicated these problems may be, however, they are, the Unexceptionalists say, hardly “unique to cyberspace.” “[I]dentical problems arise all the time in real space”;[22] the issues raised by Digitalbook’s conduct are “no more complex or challenging than similar issues presented by increasingly prevalent real-space events such as airplane crashes, mass torts, multistate insurance coverage, or multinational commercial transactions, all of which form the bread and butter of modern conflict of laws.”[23] After all, the Unexceptionalists (correctly) note, people have been communicating and transacting with other people in other territorial jurisdictions for a long time, beginning long before the Internet raised its head. Over the past century or so, a number of important principles of law and analytical tools have evolved to answer the jurisdictional problems posed by border-crossing transactions. These traditional principles and tools do not spontaneously disintegrate when confronted by cyberspace’s version of the border-crossing transaction; the algorithms that determine the applicable law(s) for any particular event – though those algorithms were, necessarily, developed to deal with realspace phenomena -- do not somehow misfire, or spontaneously decompose, when we apply them to phenomena on the global electronic network. Cyberspace transactions “are not significantly less resistant to the tools of conflict of laws than [these] other transnational transactions,”[24] and it would be a mistake to “underestimate the potential of [these] traditional legal tools and technology to resolve the multijurisdictional regulatory problems implicated by cyberspace.”[25]

Those who think otherwise – the “Exceptionalists” – err, in Goldsmith’s view, because they are “in the grip of a nineteenth century territorialist conception of how ‘real space’ is regulated and how ‘real-space’ conflicts of law are resolved.”[26] This discredited territorialist conception – “hermetic territorialism,” Goldsmith calls it[27] – involves a belief that there must be “a unique governing law for all transnational activities,”[28] a “single legitimate governing law for transborder activity . . . based on a discrete territorial contact.”[29] Hermetic Territorialism requires that we analyze Digitalbook.com’s activities to identify one body of law applicable to its behavior, to define the “discrete territorial event” which is a “necessary prerequisite to the application of local law” to Digitalbook’s conduct.[30]

Hermetic Territorialism, though it held sway for several hundred years, was “repudiated” as part of a “revolution[ ] [in] conflict of laws in the second half of [the 20th] century.”[31] Many factors – including “changes in transportation, communication, and in the scope of corporate activity [leading] to an unprecedented increase in multi-jurisdictional activity”[32] – led directly to an “expansion of the permissible bases for territorial jurisdiction.”[33] The result is that in “modern times,” a transaction “can legitimately be regulated” not only “by the jurisdiction where the transaction occurs [and] the jurisdictions where the parties burdened by the regulation are from,”[34] but also by “the jurisdictions where significant effects of the transaction are felt”[35]; under “current conceptions of territorial sovereignty,” a sovereign “is allowed to regulate extraterritorial acts that cause harmful local effects unless and until it has consented to a higher law (for example, international law or constitutional law) that specifies otherwise.”[36] If that means, as it often does mean, that more than one jurisdiction can legitimately apply its law to single events or transactions, so be it; in the modern view, “more than one jurisdiction can legitimately apply its law to the same transnational activity,”[37] and there is no need to find the “discrete territorial event” on which to base the application of any single body of law.

As a result of this change in viewpoint, extraterritorial regulation is “commonplace in the modern world”; “customary international law” and the U.S. Constitution both “permit a nation to apply its law to extraterritorial behavior with substantial local effects.”[38] It is, for instance “relatively uncontroversial”[39] that a newspaper publisher “is liable for harms caused wherever the newspaper is published or distributed.”[40] There is “nothing extraordinary or illegitimate” about this “unilateral regulation of transnational activity that affects activity and regulation in other countries.”[41] Singapore’s, or England’s, regulation of Digitalbook.com is “no less legitimate than the United States' regulation of the competitiveness of the English reinsurance market, which has worldwide effects on the availability and price of reinsurance.”[42]

The bottom line: “It is settled with respect to realspace activity”[43] – elsewhere Goldsmith refers to this as an “uncontested assumption”[44] – “that a nation's right to control events within its territory and to protect its citizens permit[ ] it to regulate the local effects of extraterritorial acts,”[45] that “prevailing concepts of territorial sovereignty permit a nation to regulate the local effects of extraterritorial conduct even if this regulation produces spillover effects in other jurisdictions, [and that] such spillover effects are a commonplace consequence of the unilateral application of any particular law to transnational activity in our increasingly interconnected world.”[46]

And if all that is “settled with respect to realspace activity,” why would we think that cyberspace is any different?

The Exceptionalists

I am, I admit, an Exceptionalist; I think it does matter, most profoundly, that Digitalbooks.com is “in cyberspace,” I think that the questions raised by its conduct are indeed different, and more difficult, than the analogous questions raised by its realspace counterpart, and I don’t think that we can resolve the jurisdictional dilemmas posed by Digitalbooks’ transactions by applying the “traditional legal tools” developed for similar problems in realspace.

I must concede, however, that the Unexceptionalist argument has a kind of syllogistic power:

(a)there are “settled principles” of “customary international law” that would apply to the activities of an ordinary, bricks-and-mortar bookstore – Analogbooks, Inc. – if it were standing in Digitalbooks’ shoes;

(b)these principles – in the “modern view,” at least – tell us that Singapore, and England, have a “legitimate interest” in regulating the realspace conduct of anyone (including Analogbooks) whose conduct produces “substantial local effects” within those jurisdictions;

(c)Digitalbooks’ transactions are “the same as” Analogbooks’ transactions, inasmuch as cyberspace transacitons are “functionally identical to transnational activity mediated by other means, such as mail or telephone or smoke signal”;[47]

(d)therefore, Singapore and England have a legitimate interest in regulating Digitalbooks’ conduct, a recognized right to regulate the conduct of persons outside its borders whose cyberspace conduct produces substantial local effects in those countries.

This syllogism, like so many others, is simultaneously both powerful and empty. The logic is unassailable: If X is true – “settled law” – in environment 1, and environment 2 is identical to environment 1, then X is true in environment 2.

But it is an odd kind of logic for these questions. Consider the notion that these questions are “settled.” As Goldsmith himself points out, the “modern view” of international jurisdiction is itself a product of profound changes in the world over the past two centuries or so. It was once “settled” law that a state could not regulate extraterritorial acts, whether or not those acts had local effects. Fifty or 100 years ago, the Unexceptionalists of the time could have pointed to the existence of these “settled” principles to suggest that the introduction of rail transport, or the telephone, or radio broadcasting, should have no effect on our analysis of jurisdictional problems. We can imagine the following colloquy:

Scene: A New York streetcorner, circa 1900. Two law professors – Professor E and Professor U – meet.

E: “Have you noticed? This telegraph thing changes everything! I can step inside a Western Union office in New York and execute a contract in San Francisco instantaneously! Incredible, eh?”

U: “Well, I suppose it is. But what of it?”

E: “What of it? Surely you jest. The world as we know it will never be the same. We’re going to need new principles of law to deal with this phenomenon. Our jurisdictional principles – especially that one that requires physical presence for the exercise of “jurisdiction to prescribe” – must yield to this new context, no?”

U: “Not at all. Transactions completed by telegraph are functionally identical to those completed by mail or by smoke signal; they all involve real people in one territorial jurisdiction either (i) transacting with real people in other territorial jurisdictions or (ii) engaging in activity in one jurisdiction that causes real-world effects in another territorial jurisdiction. It is settled law that the people of California cannot reach people and transactions occurring outside of its borders. Why would we need to adjust those principles now?”

In retrospect, of course, the Unexceptionalists were proven wrong; the world, though it may have been difficult to see at the time, was changing, and changing profoundly; over time, settled understandings became unsettled because of that change.

The world, sometimes, does that – change profoundly. When it does, settled understandings can, and do, change with it. Unless one thinks the world cannot change again, and again unsettle our settled understandings, questions about what states may or may not legitimately do can hardly rest on the notion that those questions are somehow already “settled.”

Thinking About Change

“[A]ctivity in cyberspace,” the Unexceptionalists say, is “functionally identical to transnational activity mediated by other means, such as mail or telephone or smoke signal.”

What could that possibly mean? Coming up with ways in which activity in cyberspace is obviously and incontrovertibly “functionally different” from activity in realspace doesn’t take a great deal of insight or deep thinking – in fact, it is trivially easy. In cyberspace, I can

  • instantaneously (or nearly so)
  • at zero marginal cost (or nearly so)
  • communicate an offer to sell some product or service
  • which offer can be stored, and retrieved, by the recipients at zero marginal cost
  • with zero probability (or nearly so) of error in the reproduction or distribution of that offer
  • to several million people, including hundreds of thousands in Singapore and the United Kingdom
  • each of whom has the ability to translate my message into whatever natural language he or she is most comfortable with
  • instantaneously
  • at zero marginal cost
  • and each of whom has the capacity to respond my offer
  • instantaneously
  • at zero marginal cost.

I cannot do that – that is, I cannot engage in a transaction having all of those features simultaneously – using smoke signals, telegraphs, or telephones.

The Unexceptionalists are intelligent and sophisticated thinkers; how could they possibly think that activity in communication in cyberspace is “functionally identical” – not, mind you, merely “functionally similar” or “roughly equivalent,” but identical – to realspace communication? What could possibly be going on?

The question whether activity in cyberspace is “identical to” or “different from” activity in realspace is – like its cousin, the question “is cyberspace really a ‘place’?” – a curious one. It’s like asking whether life on land is “identical to” or “different from” life in the ocean. The answer is that it is, simultaneously, both. Some principles operate identically in the two environments – natural selection, for example; others – e.g., gravity – do not. For the purpose of understanding some of the features of life in the two environments, the differences between them are irrelevant; for the purpose of understanding other features of life in the two environments, the differences are very relevant indeed. Ignoring those differences when considering questions in the latter categoryis foolish indeed.

The Unexceptionalists are undoubtedly correct: Some things – perhaps many things – about events and transactions are identical in realspace and in cyberspace. Plus ca change, plus ca le meme chose, nothing new under the sun, and all that. Transactions between human beings, whether via e-mail or postcards or smoke signals, are still transactions between human beings, and they are, therefore, necessarily the same in many respects. Whatever it is that motivates human beings to do one thing or another – love, hate, greed, curiosity, fear, and the rest – has not changed with the coming of the Internet. Digitalbooks.com, just like Analogbooks, is providing a forum where buyers and sellers can exchange consideration for goods; Digitalbooks.com, like Analogbooks, has to have a system for making sure that goods get shipped from seller to buyer after a transaction is consummated; Digitalbooks.com, like Analogbooks, has to have rules for identifying the winners and losers of individual auctions; Digitalbooks.com, like Analogbooks, has to have means for obtaining payment for its services, accounting for those payments, and transferring money to its suppliers.