As Boundaries Fade: The
Social Contract in Cyberspace
By Steven J. Horowitz,
For over ten years, scholars have debated over law in cyberspace. Some,
the “exceptionalists,” argue that cyberspace should develop its own system of
laws and regulations. Others, the “unexceptionalists,” argue that real-space
territorial law must govern the internet. This paper advocates a new kind of
exceptionalism, grounded in an examination of legitimate authority in
cyberspace. I use social contract theory to locate two sources of legitimate
cyberspace authority: the authority of a real-space sovereign over its citizens
and the authority of a cyberspace community formed by a distinct social
contract. I argue that, because cyberspace dissolves territorial boundaries,
internet users are insecure in their knowledge of political relationships and
that cyberspace communities can resolve this inconvenience.
I. Framing the Question
Cyberspace is a world without territorial boundaries. As we
travel through cyberspace, we have no way of knowing where our transactions
take place. Since law tends to correspond to territorial boundaries, cyberspace
calls into question the legitimacy of real-space law on the internet. This
paper aims to answer the question of legitimacy: what kind of legitimate
authority exists in cyberspace? I use social contract theory to identify two
distinct sources of legitimate authority on the internet: 1) the sovereignty of
a real-space state over its citizens and 2) the power of a cyberspace sovereign
to govern the default laws of a bordered part of cyberspace. The power of a
real-space sovereign over its citizens does not disappear as they enter
cyberspace, since this power is derived from a social contract that does not
depend on geography. But internet users can still form distinct communities in
cyberspace, built on social contracts. Such cyberspace communities provide security
for internet users. Without a distinct social contract in cyberspace, one is
left to wonder whether those with whom he interacts will follow a common set of
rules and be subject to a common authority. But as we form cyberspace
communities, built upon social contracts, we are able to interact with other
users, knowing that we are in a civil society.
Before engaging the theoretical questions
central to this paper, I present two hypothetical situations. Each involves a
conflict that lacks a clear resolution because it is unclear where legitimate
authority might lie. While I intend to return to these later in order to
provide some solutions, I cannot do so until I build a theoretical framework.
The first example involves an online
auction. Suppose a seller located in country X sells a given product on his
website through an online auction. The winning bidder lives in country Y. The
buyer provides his credit card number in order to pay for the product, but the
seller charges him an “international service fee” of $200. Assume that, in
country Y, imposing such a fee without prior consent would be illegal, whereas
in country X, it would not be. If the buyer could not cancel his transaction,
how could he resolve this problem? The seller is not doing anything illegal
insofar as he is not transgressing any law that he lives under. The seller is
dealing with a buyer under a different set of laws, however. So whose laws
should prevail? In practice, there are various factors that would influence the
outcome, some more just than others. But since such disputes can easily arise
within cyberspace, and since our current system may or may not provide a just
resolution, we should consider what the possibilities might be and whose
authority is legitimate.
Consider a second example. In this case,
the people involved are two players in a virtual gaming environment. Within
this environment, there are goods that players can acquire through trade in
virtual currency or through successful battles with virtual monsters. Suppose
that player A owns a powerful sword, earned by defeating the most imposing
beast in the game. One day, player A, within the virtual world, puts the sword
down for a moment in order to free his hands to do something else. Player B,
who is from another real-space country than player A,
sees the sword on the ground next to player A and takes it. One could argue
that player B does not have the right to use the sword because A “owns” it
somehow, or, conversely, one could argue that there is nothing that should stop
B from picking up goods that are on the ground in the virtual environment. But
if player A wishes to get his sword back, to whom
could he appeal? In this case, it is likely that neither country’s laws extend
in any obvious way to swords in virtual gaming environments. But even if each
country had applicable laws, whose laws should prevail?
Where is there a legitimate authority to decide who should get the sword? The
answers are unclear.
What these examples demonstrate is that
often there is no clear source of legitimate authority on the internet. Even
when one country’s laws prevail, it is unclear whether they do so because they
legitimately ought to or because that country has greater political power than
its counterpart. The internet itself does not appear to provide a legal
framework of its own to resolve disputes, and even if it did, it is unclear whether
such a framework would be legitimate either. The solution to these problems
depends in part on the source—if any exists—of legitimate sovereignty on the
internet. The rest of this paper is a search for this legitimate political
power.
In order to locate such power, I examine
the social contract in cyberspace. In section II, I discuss the two prevailing
positions in the cyberspace governance debate. In section III, I provide the
social contract theory that will frame my analysis. In section IV, I analyze
the social contract in cyberspace. In section V, I discuss the implications of
the social contract in cyberspace for the governance debate, presenting a new
kind of exceptionalism.
II. The
Governance Debate
For more than ten years, scholars have debated
over the law of cyberspace.1 One can divide
the debate into two schools of thought: the exceptionalists and the
unexceptionalists.2 Exceptionalists argue that cyberspace cannot and should not
be governed by real-space law. Unexceptionalists, on the other hand, argue that
cyberspace is not uniquely difficult to regulate and that real-space law can
regulate cyberspace effectively. At first, these views might appear
diametrically opposed. The unexceptionalists, however, attack only the
infeasibility of territorial law in cyberspace without addressing the normative
claim that territorial law should not govern cyberspace. These schools
of thought therefore end up talking past each other: the unexceptionalists are
tied up in descriptive claims while the exceptionalists focus on normative
ones.
A. The
Exceptionalists
If the exceptionalists have a manifesto,
it is John Perry Barlow’s famous “Declaration of the Independence of
Cyberspace.”3 Barlow begins his declaration in the following way:
Governments of the Industrial World, you
weary giants of flesh and steel, I come from Cyberspace, the new home of Mind.
On behalf of the future, I ask you of the past to leave us alone. You are not
welcome among us. You have no sovereignty where we gather.4
This declaration sets the tone for the
exceptionalist platform. Barlow argues that 1) there is no social contract
between citizens of cyberspace and real-space governments,5 2) cyberspace
will create its own laws,6 and 3) real-space governments could not
govern cyberspace even if they wanted to. Barlow envisions a democracy in
cyberspace better than any real-space counterpart, one unencumbered by the
shackles of materiality. While Barlow’s manifesto is eloquent and inspiring,
his argument is not fully developed.
The best scholarly articulation of
Barlow’s vision is that of David Johnson and David Post.7 Johnson and Post
argue that real-space law is not legitimate in cyberspace. I will focus on the
two most important reasons why Johnson and Post think that cyberspace should be
left to govern itself. First, cyberspace dissolves traditional notions of
boundaries. Second, cyberspace makes providing notice for users as to what laws
apply to their transactions difficult.
Johnson and Post argue that “Cyberspace
has no territorially based boundaries.”8 Because cyberspace exists
everywhere simultaneously, one cannot draw up geographical boundaries within
its spaces. There are only two boundaries that one can reliably discern. The
first is the boundary between real-space and cyberspace: cyberspace is “inside”
the internet while real-space is “outside.” One can easily tell whether she is
in cyberspace or real-space.9 The
second is the boundary that separates distinct spaces within cyberspace. For
example, one knows whether she is in “Second Life”10 or in the
blogosphere.11 What has important
implications for the appropriateness of real-space law in cyberspace, however,
is that 1) it is difficult and potentially expensive to locate where, in
real-space, each individual is12 and 2) the “space” in which
cyberspace transactions take place is impossible to pinpoint.13
This lack of boundaries calls into
question the legitimacy of territorial law in cyberspace because law, according
to the exceptionalists, is fundamentally tied to boundaries.14 Territorial law
is based on territories, which are defined by their boundaries. Johnson
and Post use trademarks as an example of how important boundaries are for law,
citing that, “In the United States, similar businesses can even use the same
name, provided there is sufficient geographic separation of use to avoid
confusion.”15 But cyberspace does not allow for geographic
separation because there are no geographic boundaries. Territorial trademark
law, as it operates in real-space, does not make sense in cyberspace.
Not only do we not know where in
real-space our cyberspace activities take place, but we also cannot know for
certain where the effects of our cyberspace activities are felt. Johnson and
Post explain that “the effects of online activities [are not] tied to
geographically proximate locations.”16 This means that the real-space
distance between persons in cyberspace has no bearing on their ability to
affect one another. The effects of cyberspace actions can be felt anywhere in real-space
at any time. One way to think about law is that it regulates actions to
minimize negative effects. But different legal systems do not always agree as
to which effects are negative. So, as we act in cyberspace, not knowing where
the (potentially negative) effects of our actions are felt, we cannot be sure
whether we are unknowingly breaking the law abroad.
This problem is a problem of notice.
Cyberspace is unable to provide notice as to where, in real-space, transactions
take place and what laws apply. In real-space, boundaries provide notice of
what rules govern activities within a given space. When one enters
Exceptionalists argue that subjecting
people to territorially based law in cyberspace is unfair. Cyberspace has no
territorially-based boundaries, one can never know where the effects of one’s
actions are felt, and cyberspace can provide no notice to those who cross
real-space boundaries while surfing the internet. For these reasons,
exceptionalists argue that real-space law is illegitimate in cyberspace. They
suggest that cyberspace, or distinct spaces within cyberspaces, can and will
provide their own regulatory mechanisms.19
One problem with this kind of exceptionalism
is that it depends on a given structure of cyberspace. But what if we could
redesign cyberspace to provide territorial notice? What if we could engineer a
cyberspace that conformed to territorial borders?20 This is one problem with
deriving a normative claim from a descriptive one—things can change.21 Johnson and Post
are actually committed to changes in cyberspace in that they believe regulatory
structures will emerge in cyberspace.22 What the exceptionalists need is
a normative explanation as to why their vision of cyberspace’s future ought to
be realized and/or some evidence as to why it is likely.23
B. The
Unexceptionalists
Unexceptionalists argue that, while
cyberspace may present some new problems for law, there is nothing particularly
special about law in cyberspace. Real-space law can provide an effective means
of regulating activities on the internet. Jack Goldsmith, who has made the
strongest arguments against exceptionalism, says that Johnson and Post make
three mistakes: making cyberspace seem more novel than it is, ignoring the
distinction between default and mandatory laws, and understating the potential
effectiveness of technology and traditional legal tools for dealing with
problems.24
Goldsmith and the unexceptionalists argue
that transactions across territorial boundaries in cyberspace are not all that
different from transactions across territorial boundaries in real-space:
Goldsmith calls them “functionally identical.”25 In each case, a
person in one state interacts with a person in another state. Where there is
harm, the state in which the person harmed is located might seek to exact
punishment. We could tell the same story about e-commerce or international mail
orders. Though there may be more transactions across borders because the
internet provides a more effective medium, there is nothing special about how
the law should handle cyberspace. Judge Easterbrook goes so far as to liken the
exceptionalists’ “law of cyberspace” to a “law of the horse.”26 He argues that
cyberspace is one of many things that law regulates, and “the best way to learn
the law applicable to specialized endeavors is to study general rules.”27
David Post argues that Goldsmith is wrong
about cyberspace transactions being “functionally identical” to certain
real-space transactions.28 He cites six characteristics of
cyberspace transactions that cannot apply to real-space transactions. In
cyberspace, transactions can occur:
instantaneously (or nearly so); at zero
marginal cost (or nearly so); to several million people; with near-zero
probability of error in the reproduction or distribution of that offer; which
can be stored, retrieved, and translated into another language by each of the
recipients (instantaneously, and at zero marginal cost); and to recipients who
have the capability to respond to my offer (instantaneously, and at zero
marginal cost).29
Given these unique characteristics, it is
hard to identify what real-space transactions could possibly be “functionally
identical” to cyberspace transactions. And, if this is true, it is equally hard
to identify what existing laws we should use to regulate cyberspace
transactions.
Goldsmith’s second argument against the
exceptionalists is that they “do not attend to the distinction between default
laws and mandatory laws.”30 Private parties are free to adapt
default laws as they please. So, for example, Major League Baseball can decide
that it is acceptable to spit while playing baseball but not to run out of the
baseline. Mandatory laws, however, cannot be so tailored. Even if Major League
Baseball should want to allow players to smoke marijuana, the league could not
do so. The laws that prohibit possession and use of illegal substances are
mandatory—they are not something that a private entity can decide to turn off.
In the same way, we can develop law in cyberspace, but it can affect only
default law: we can ban spamming in certain settings since there is no
mandatory “right to spam.” But the law of cyberspace cannot abnegate mandatory
laws, and therefore, the law of cyberspace is not distinct from real-space law.
This criticism of exceptionalism is not a
strong one. It strengthens the exceptionalist position insofar as it allows for
the legitimacy of self-regulatory structures in cyberspace, at least for default
laws. Are the exceptionalists right about the difficulties of applying
territorial mandatory law in cyberspace? Since different states have different
sets of mandatory laws, and since cyberspace provides no better notice for
mandatory law than for default law, the exceptionalists’ problems remain. So
the distinction between default and mandatory law does not take us far. 31 The
exceptionalists, by not attending to this distinction, have perhaps lost a free
pass with respect to default law in cyberspace, but their position is not
weakened.
Goldsmith’s third argument against the
exceptionalists is that they “underestimate the potential of traditional legal
tools and technology to resolve the multijurisdictional regulatory problems
implicated by cyberspace.”32 Since the multijurisdictional problems
of cyberspace are no different from those of real-space, the tools we use for
real-space conflicts of law should suffice in cyberspace. There is no reason to
invent new tools for an old problem when our old ones are effective. In
cyberspace, we also can resolve problems by creating innovative technologies to
curtail unwanted activities so that conflicts of law are mitigated.
One can summarize these three attacks on
exceptionalism in this way: unexceptionalists think that there is no compelling
reason to treat cyberspace differently than we treat any other medium.
Multijurisdictional problems may arise, but we have established methods for
solving such problems. Perhaps new sets of laws will develop in cyberspace, but
these laws can only be default laws, and private organizations have created
default laws for some time. Cyberspace cannot and will not contravene mandatory
laws.
C. The Point at
Issue
If the unexceptionalists are right, then
one might think that this debate is fundamentally about the ability of
real-space law to deal with the problems of cyberspace. If this is the case,
then the exceptionalist argument must be that real-space law cannot govern in
cyberspace. The unexceptionalist’s burden, then, would be simply to describe
how it is possible that real-space law could resolve the seemingly complex
problems of cyberspace. I think that this is how the unexceptionalists tend to
approach the debate. They think that, if they can prove that real-space law can
adapt to the problems of cyberspace, then they have disposed of the need to
address the exceptionalists any further: the exceptionalists become romantic
cyberanarchists and are easily dismissed.33
The unexceptionalists are wrong about the
exceptionalist position, however. It is not enough to address feasibility
because the exceptionalists are not interested in what is possible but what is
right. Exceptionalist arguments might sound descriptive: that there are no
territorial boundaries in cyberspace is a descriptive claim. But their deeper
concern is normative. Because there are no territorial boundaries in
cyberspace, and because borders are necessary for the legitimacy of territorial
law, we ought not to subject citizens of cyberspace to territorial law.
While the unexceptionalists do a fine job of offering ways in which we could
apply real-space law to transactions in cyberspace, they do not address the
core normative commitments of exceptionalism directly.
Exceptionalists tacitly agree that we
could treat cyberspace law as if it were no different from real-space law and
try to work with the structures we already have in place. It just wouldn’t be
right. The unexceptionalists are focused on what is possible while the
exceptionalists care about what ought to be.
The normative question that began the
cyberspace governance debate is this: what laws should govern cyberspace? The
exceptionalists argue that internet users should develop communities that provide
a set of rules for a given space—a law of cyberspace. The unexceptionalists are
committed to the feasibility of regulating cyberspace with real-space law. The
laws that should govern cyberspace are those that have legitimate authority to
do so. In order to locate such legitimate authority in cyberspace, I now turn
to social contract theory.
III. Social
Contract Theory
In this section, I provide some
foundation in social contract theory upon which I will build my arguments
concerning the political state of cyberspace. Though I will be drawing from
John Locke’s version of the social contract in his Two Treatises, most
of what I offer is not unique to Locke.
John Locke argues that the social
contract is the necessary and sufficient condition for civil society. With it,
there is a political body, joined together under a common government. Without
it, there is no common authority. He explains:
The
only way whereby any one divests himself of his Natural Liberty, and puts on
the bonds of Civil Society is by agreeing with other Men to joyn and unite
into a Community…When any number of Men have so consented to make one
Community…they are thereby presently incorporated, and make one Body
Politick.34
Thus,
the formation of society requires consent: this consent represents the
social contract. The social contract is an agreement to give up certain natural
freedoms and to live under the laws of society in exchange for protection.
Locke provides a simple test for whether
such a social contract exists in a given situation. To determine whether a
group of individuals are in civil society, all one must do is find out whether
they have a common authority to appeal to or a common set of laws: “Those who
are united into one Body, and have a common establish’d Law and Judicature to
appeal to, with Authority to decide Controversies between them…are in Civil
Society.”35 If Paul and Chuck get into a fight over who owns a
given book, for example, they are in civil society if there is some common
authority to which they can appeal to resolve the dispute. But what if they do
not have such an authority? Locke explains, “Those who have no such common
Appeal…are still in the State of
On Locke’s account, the state of nature
is no mere metaphor.37 It is a normative relationship
between parties. If two parties cannot appeal to a common authority, then they
are in a state of nature “in reference to one another.”38 This is what I
call the relativity of the state of nature: the state of nature is
neither a fairytale nor an absolute state. It depends on the connections
between people. For example, one could argue that parties transacting through
international commerce are in the state of nature in reference to one another:
while companies may do business across national boundaries, they do not always
agree to recognize a single authority. In case of dispute, each company may
disagree as to which governing body has the right to provide a resolution. This
is not to say that international commerce is chaotic. But the lack of a recognized
common authority can complicate things.39 Were
there a common authority, consented to by the entire business world, conflict
resolution might be much easier. As it is, international traders who have no
common authority to appeal to are in the state of nature in reference to one
another.
The state of nature has its advantages.
First, while in the state of nature, people are in a “State of perfect Freedom
to order their Actions and dispose of their Possessions, and Persons as they
think fit, within the bounds of the Law of Nature.”40 The state of
nature is also a state of equality.41 Everyone has the same rights,
which means that no one can claim a natural right to resources held in common42 or to command
others as an absolute ruler.43
Since the state of nature affords perfect
freedom and perfect equality—both of which are diminished in society—one might
wonder why anyone would ever want to enter into a social contract. The social
contract and the society that emerges out of it provide one significant
advantage over the state of nature: protection. In the state of nature,
disputes quickly turn into a state of war. In civil society, a common authority
can help to resolve this state of war.44 In the state
of nature, however, the state of war is not so easy to end. If a state of
nature existed without the threat of war, then those within that state might
not choose to enter into civil society.45
The social contract is what distinguishes
civil society from the state of nature. Without it, one is in the state of
nature, enjoying freedom and equality but suffering from lack of protection
that society provides. Usually, the inconveniences of the state of nature are
enough to make us enter into civil society. Civil society provides laws, common
authority, and a judiciary, but it requires that each citizen give up some of
her natural freedoms.
IV. Social
Contracts in Cyberspace
Social contracts in real-space tend to
correspond to geographical boundaries. People living within a given territory
agree to abide by the sovereign’s laws in exchange for protection under those laws.46 By dissolving territorial
boundaries, cyberspace clouds our ability to identify political states between
parties. In this section, I discuss the political state of cyberspace in two
parts. First, I look for the social contract in cyberspace.47 Second, I discuss the implications of the social contract on
the political state of cyberspace.
A. Finding the
Social Contract in Cyberspace
Locke’s simple test for the existence of
a civil society should determine whether cyberspace is a state of nature or a
civil society. If the citizens of cyberspace can resolve disputes by appealing
to a common authority and they have a common set of laws that they agree to
abide, then they are in a civil society. In other words, if there is a social
contract, there is a society. Without a social contract, there is only the
state of nature.
To subject cyberspace to this test, let
us consider an example. Bob creates a website in country X with some
interesting images that he creates himself. Tom, in country Y, sees the site
and likes the images. Tom copies the images and uses them on his own site.
Assume that X’s copyright laws would prohibit such copying but that country Y’s
laws would not. If Bob could not convince Tom to take the images down, to whom
could Bob appeal? Bob could ask country X to provide force to protect his
intellectual property interests, but that would not be an appeal to a common
authority. There is no common authority between Bob and Tom. Perhaps they
or their countries could forge reciprocal agreements. But this would still not
constitute an appeal to a common authority. The reciprocal agreements in
question would not allow Bob and Tom to appeal to a common authority because
each would have to appeal to his own sovereign. But if the
sovereigns were to disagree as to how to resolve the situation, then Bob and
Tom could go no further. This is because a reciprocal agreement is a
contract, but not necessarily a social contract, which would provide a
common authority. The only type of agreement that counts as a social contract
is “one of agreeing together mutually to enter one Community, and make one Body
Politick.”48 Therefore, Bob and Tom would remain in the state of
nature in reference to one another.
Would this be any different if Bob and
Tom were both citizens of country X? Yes. In that case, Bob and Tom would each
have agreed to the laws of country X, which prohibit such copying. If Bob wanted
to take legal action against Tom, the courts of country X would have the
legitimate authority to resolve the dispute and exact punishment, if necessary.
Bob and Tom would share a common social contract, and they would be in civil
society. Therefore, cyberspace is not exclusively a state of nature. Instead,
those citizens of cyberspace that are also citizens of the same real-space
country are not in the state of nature in reference to one another.
It might seem as though the opposite is
also true: people in cyberspace who are not citizens of the same real-space
country are in the state of nature in reference to one another. But this is not
necessarily the case.
Virtual worlds provide a context in which
citizens of distinct real-space states can share a common social contract. For
example, what if Bob and Tom fight over virtual property in a virtual world?49 Assume
that the residents of this virtual world have created a virtual government that
has the legitimate authority, gained from the consent of the users it governs,
to resolve disputes over property and that Bob and Tom both have agreed to live
under that government. If that were the case, then Bob would be able to appeal
to the authority of the virtual government to help resolve his dispute because
Bob and Tom would be in civil society in reference to one another.
This virtual world example helps
illuminate an interesting aspect of the social contract in cyberspace: internet
users can form distinct social contracts to create legitimate cyberspace authority.50 As long as those
interacting in a given place on the internet agree to give up some rights to
some governing body in charge of resolving disputes, there is a social
contract. Users of Ebay, for example, agree to follow a certain set of rules.51 In return, they
receive Ebay’s protection: Ebay will keep track of feedback and expel users who
do not obey its rules. This exchange constitutes a kind of social contract, and
therefore users of Ebay are in a kind of civil society.
To this, one might object that the civil
society of the Ebay community is illusory, or that its social contract is
useless without the support of the
There are two responses to this
objection. First, while it is true that Ebay operates under real-space laws and
uses real-space law to enforce its own rules, that
does not mean that cyberspace authority always relies on real-space
governments. An online community can create its own enforcement mechanisms
without appealing to real-space law. Here is a simple example. Suppose I create
a virtual world that anyone can enter, as long as they agree not to fight with
one another. Those who come in must enter into a social contract with me and
can enjoy the benefits of my world, whatever they may be, as long as they abide
by this law. There is only one punishment: exile. If you break the law, you are
no longer permitted in my virtual world. Social contracts in my world do not
require an appeal to any authority other than whatever authority exists within
my world.53 This example shows that,
while online entities may choose to rely on the support of real-space
sovereigns, they need not do so.
The second response to this line of
criticism is that it is irrelevant. The question is not whether an online
authority could exist without the support of a real-space authority. The
question is, instead, whether a legitimate online authority can exist that 1)
is unique to cyberspace and 2) derives its authority to govern its part of
cyberspace from the consent, through a social contract, of the community of
users in that part of cyberspace. The authority to govern (presumably
international) users in a distinct part of cyberspace will not come from the
support of a real-space sovereign, even if the online authority depends on a
real-space sovereign—it will come from the consent of the governed. One could
argue, for example, that
Still, there might seem to be a conflict
between online and real-world social contracts. One might argue that committing
oneself to different social contracts is a problem, but we do that all the
time. We agree to follow the laws of a country, but we also agree to a
different set of laws within our state or province. Social and political
organizations also provide a set of benefits in exchange for some freedom.54 The only restraint on our ability to enter into different
social contracts is that these contracts must not conflict with the mandatory
laws of the state. For example, if an Australian has agreed not to steal from a
local store under the laws of the state, he cannot join an organization that
requires that its members do so without breaking his social contract with the
state.
This is where the distinction between
default and mandatory laws can become important.55 One cannot enter
into a cyberspace social contract that contravenes the mandatory laws of a
sovereign state with which he has previously entered into a social contract.
So, to use a well-known example, if the
One can, however, enter into a social
contract with any sort of default laws that she wishes to respect. In this
case, she still upholds her social contract with her real-space sovereign.
There are three relevant possibilities
for the social contract between agents in cyberspace. First, two agents can
have no common social contract at all. In this case, the two agents are
citizens of distinct territorial sovereigns and have not entered into a common
social contract in cyberspace. There is no common authority to resolve their
disputes, and they are in a state of nature in reference to one another.
Second, two agents can have identical social contracts. They would then be able
to appeal to their common territorial sovereign and/or their common cyberspace
sovereign, if one exists. Third, two agents can have different territorial
commitments but identical cyberspace commitments. They could appeal to the
cyberspace sovereign to resolve those disputes that the cyberspace community,
through the social contract, has given the sovereign the authority to resolve.
B. The Political
States of Cyberspace
The political state of cyberspace depends
on the relationships between agents there. Because there are several
possibilities for the social contract between agents, cyberspace cannot have a singular
political state. Many scholars argue that cyberspace is a kind of
undeveloped frontier, reminiscent of a Lockean state
of nature.56 If this view is correct, then the only law of
cyberspace is natural law. But many people in cyberspace share common social
contracts. For example, there are over 120 million Chinese internet users, and
every one of them lives under the authority of the Chinese government.57 It makes little sense to say that all of them are in the
state of nature in reference to each other. On the other hand, cyberspace
allows for interactions between people living under different sovereign states,
respecting different laws. If two such people were to interact, they would be
in the state of nature unless they were members of a common cyberspace
community. So, in general, we cannot say that cyberspace is either a state of
nature of a civil society. One can only make such determinations on a
case-by-case basis by examining the relationships in play. Yet there are still some
general observations that one can make concerning the political states of
cyberspace.
As in real-space, the default political
relationship between agents in cyberspace is the state of nature. Only if two
agents enter into a social contract with each other or with the same political
body is their political relationship one of civil society. So an Australian in
What is distinct about cyberspace,
however, is that almost everyone could be a stranger in a strange place. In
real-space, while the default political relationship is the state of nature,
people who live within the boundaries of a sovereign state are usually in civil
society, having entered into a social contract with that sovereign. And most of
the time, we stay within the boundaries of the sovereign state. We tend to
assume that our transactions are governed by a common authority and that we are
in civil society. Since cyberspace does not conform to traditional territorial
boundaries, we rarely know whether someone we meet is part of our territorial
society. This is the political challenge of cyberspace.
One solution to this challenge is to
enter only those places in cyberspace that require a social contract. This
ensures that everyone within a given space respects at least some of the same
laws. One can even decide which rules matter most to him and then build a
virtual world for those who wish to follow those rules. Borders provide
security that a community of users who agree to follow the same rules. These
cyberspace borders can be more effective than those of real-space because
technology is often a more effective gatekeeper than the fences and guards of
real-space.59 But this solution is
imperfect. There will be places worth visiting that do not provide the
protection of an online social contract. In these places, one must choose
whether to bear the inconvenience of the state of nature or forgo visiting.
To summarize, the political state of
cyberspace is amorphous. Cyberspace provides a vast medium for the interaction
between various people in various political relationships. Sometimes, people
are in civil society because of their common real-space political commitments.
At other times, the only way to avoid the state of nature is to enter into a
social contract in a given place within cyberspace.60 Cyberspace
provides a political melting pot, bringing together anyone from anywhere to
forge whatever political relationships they will.
V. Conclusion
In this final section, I address the
implications of the political states of cyberspace both for how we ought to
analyze legal and political problems and for how we ought to resolve the
cyberspace governance debate. I argue that social contract theory teaches us
that we should consider not territorial boundaries but the relationships
between people to identify legitimate political power. I further argue that
cyberspace is unique in that it challenges our real-space assumption that most
people we meet will be members of our sovereign state.
A. People, not Places
Those involved in the governance debate
examine real-space law across boundaries in cyberspace.61 The exceptionalists argue that applying territorially-based
law in cyberspace is inappropriate. The unexceptionalists argue that it is
possible (perhaps desirable). But each position commits itself to the
assumption that places are the most significant factor in determining the
legitimacy of law.62
We have seen, however, that one way to
think of legitimacy is through the social contract. The social contract is a
mutual agreement between people. In real-space, there is a necessary
connection between people and places, so the social contract within a set of
boundaries is usually identical to the social contract between people in that
set of boundaries. But since cyberspace does not respect real-space boundaries,
the sometime identity between people and places disappears. The more
significant element in political power is people: if the people of cyberspace
agree to respect a common authority, then they are in civil society
irrespective of their locations.
It is tempting to proclaim that
cyberspace is a state of nature (or was, some time ago). This kind of sentiment
is behind Barlow’s Declaration.63 But the state of nature does not
describe places, either. The agreements between people determine the political
state of cyberspace. Many people, inevitably, will have common real-space
agreements before they step into cyberspace. They cannot abandon these
commitments simply by surfing the internet. These people, at least, are in
civil society. We have also seen that people can form social contracts distinct
to cyberspace, but these cannot ignore the mandatory laws of real-space
sovereigns.
One cannot ignore that cyberspace really
is different from any previous medium, however. The primary difference is a
difference of scale.64 The number of possible interactions in cyberspace is
almost without limit, as there are roughly one billion internet users in the
world,65
and the internet makes the distances between these users disappear. This should
not affect how real-space sovereigns regulate the activities of their own
citizens, but it does allow for disputes between many people who are in the
state of nature in reference to each other.
As such disputes arise, we may be tempted
to use force to establish our claims.66 But such
use of force should not be taken lightly. If an individual (or a sovereign
acting for him) uses force against another in the state of nature, that begins
the state of war. Locke explains, “Force…where there is no common
I think that the exceptionalists are
right to worry. It is so difficult to difficult to determine the political
states between parties within cyberspace that the legitimacy of real-space law
is often questionable. If we can forge distinct social contracts in cyberspace,
then at least we can protect ourselves from certain activities within a given
space. These social contracts cannot go against the mandatory laws of our
real-space sovereigns, but that does not imply that they won’t be useful or
effective. Since real-space law has not yet caught up to the problems unique to
cyberspace, the laws of cyberspace can at least provide protection from our
newest demons.
B. Exceptionalism Revisited
The completed discussion of legitimate political
power in cyberspace brings us back to the question of cyberspace governance. In
this final section, I explain why social contract theory strengthens a kind of
exceptionalism. Like the exceptionalists, I argue that (1) there can be
legitimate cyberspace civil societies that do not derive their legitimacy from
territorial sovereigns. Further, I argue (2) these distinct authorities are
desirable. Unlike the exceptionalists, however, I argue that (3) the power of a
real-space sovereign to govern the cyberspace actions of its citizens is a
significant factor in cyberspace governance. Finally, I argue that (4) despite
its inherent uncertainties, there is good reason to want to retain the current
form of cyberspace. I will consider each of these conclusions in turn.
There can be legitimate cyberspace civil
societies that do not derive their legitimacy from territorial sovereigns. The social
contract is the source of legitimate authority and the necessary and sufficient
condition for civil society. If users interacting in cyberspace enter into a
community in which each user gives up some of his rights and agrees to
recognize a common authority, then, by doing so, each user legitimates such an
authority while forming a civil society. Since civil society and the authority
it creates require nothing more than the social contract for their legitimacy,
there is no need for the support of a real-space sovereign. Therefore, there
can be distinct civil societies in cyberspace.
Such distinct cyberspace authorities are
desirable. According to Locke, humans formed the first civil societies
not because civil society is inherently more desirable than the state of nature
but because the state of nature was fraught with uncertainty.68 Though
cyberspace is not a state of nature,69 we still experience a similar
uncertainty. In real-space, we are confident that, while traveling within our
own sovereign state, each person we interact with will be living under the
common authority of our sovereign state. But such confidence disappears in
cyberspace because the security that territorial boundaries provide does not
exist. We cannot know for sure whether each person with whom we interact is in
civil society with us.
The solution for such uncertainty is the
formation of distinct cyberspace authorities. If we are users in a virtual
world, for example, we want to be sure that each person we meet plays by the
same rules that we do. Since we will not know in advance whether each person we
meet is a citizen of the same territorial sovereign, one effective way to
ensure a common authority is to create one. By creating a distinct cyberspace
civil society, we create a common authority that provides security and the
knowledge that we do not need to worry about the inconveniences of the state of
nature when we meet strangers in the virtual world. Therefore, distinct
cyberspace authority is not only possible, but also desirable.
This, in fact, is the solution to each of
the problems raised in the introduction above. The dispute over an online auction
would have been more easily solved were the parties both members of the same
online community. In that case, there would be a source of authority that could
resolve the dispute, and its authority would be legitimately binding upon each
of the parties. The same goes for the virtual world example. If the sword
“theft” occurred within a world where each player has entered into a common
online social contract, then the government of that virtual world would have
the authority to resolve the dispute.
Still, the power of a real-space
sovereign to govern the cyberspace actions of its citizens is a significant
factor in cyberspace governance. If John Perry Barlow were in charge, real-
space law would not govern any cyberspace activities—he argues that real-space
sovereigns have no moral right to govern the activities of citizens of
cyberspace.70 But he is wrong for two
reasons. First, his condemnation of real-space sovereignty in cyberspace
presupposes a separation between real-space and cyberspace that does not exist.
All “citizens of cyberspace” must also be citizens of real-space. Since they
are citizens of real-space, they can be subject to real-space law. This is the
second reason that Barlow is wrong: real-space sovereigns do have a legitimate
moral right, which comes from the social contract, to govern the actions of
those cyberspace citizens who are citizens of the sovereign in real-space. If
Bob, a citizen of cyberspace, is also a citizen of
We should be glad that the moral right of
real-space sovereigns to govern their citizens extends into cyberspace. It
makes dispute resolution even easier. As mentioned above, Ebay, for example,
provides protection for its users through feedback mechanisms and the potential
for exile. But suppose person A paid person B ten thousand dollars on Ebay for
a painting that turned out to be a fake. If A and B
were both
This security of legitimate territorial
authority in cyberspace only exists because A and B
were both citizens of the same real-space sovereign—territorial law does not
legitimately extend to disputes between parties in the state of nature. While
the
Finally, despite its inherent
uncertainties, there is good reason to want to retain the current form of
cyberspace. One might wonder, if cyberspace as we know it cannot provide
sufficient notice as to what territorial laws could apply to a given
transactions, why we shouldn’t change the architecture of the internet. There
must be some imaginable form of the internet that allows for such notice. We
could even build an internet in which each user has no choice but to advertise
his real-space location at all times. But such gains would come at the cost of
the freedom that the internet offers. Internet users value their privacy, but
their privacy would be severely diminished if they should be required to
advertise their locations. While the “frontier” metaphors that we commonly use
to describe the internet may not be accurate,71 the spirit of freedom that
underlies them is important. We could have an internet with clearer territorial
boundaries, but it would mean the end of cyberspace as we know it. The cost of
such boundaries may be too high.
The exceptionalists are right about the importance of civil societies in cyberspace, but they have not isolated what is most exceptional about cyberspace. What makes cyberspace so different from real-space, and what makes civil societies in cyberspace so valuable, is uncertainty. Our real-space borders provide a degree of certainty that the state of nature is a distant memory. But territorial borders disappear in cyberspace, and the person next door in a virtual world could be half a world away in real-space. This uncertainty can be inconvenient and threatening. Distinct “laws of cyberspace,” deriving their authority from the consent of those they govern can remedy this uncertainty. The social contract can resolve the tensions that grow as territorial boundaries fade.◦
©Steve Horowitz, 2006
Notes
∗ The author thanks Robert Guay, David Post, Paul Taylor,
Nathaniel Fleischer, and Sarah Van Doel for their comments on previous drafts.
All remaining errors are his own.
1
See Jack L. Goldsmith, Against Cyberanarchy, 65 U. CHI. L. REV. 1199
(1998); David R. Johnson & David G. Post, Law and Borders—The Rise of Law in Cyberspace, 48 STAN. L. REV. 1367
(1996) [hereinafter Law and Borders]; David G. Post, Against “Against
Cyberanarchy”, 17
articles/post.shtml; David G. Post, Governing
Cyberspace, 43 WAYNE L. REV. 155 (1996).
2
I take the names from Post, Against “Against Cyberanarchy”, supra note
1.
3 John Perry Barlow, A Declaration of the
Independence of Cyberspace (Feb. 8, 1996)
http://homes.eff.org/~barlow/Declaration-Final.html. But see Neil
Weinstock Netanel, Cyberspace Self-Governance: A
Skeptical View from Liberal Democratic Theory, 88 CALIF. L. REV. 395 (2000)
(arguing that Barlow’s vision creates a less democratic space)
4
Barlow, supra note 3.
5
6
Capitanchik & Whine suggest that cyberspace has a natural tendency for
self-regulation. See David Capitanchik & Michael Whine, The
Governance of Cyberspace: Racism on the Internet, in LIBERATING
CYBERSPACE: CIVIL LIBERTIES, HUMAN RIGHTS, AND THE INTERNET 237, 245 (Liberty
ed., 1999) (“[T]he Internet relies on the good sense of users to deal with
those who ‘abuse’ its self-proclaimed ethos for their unacceptable ends.”).
7
Johnson & Post have several good articles on this, but the seminal one is Law
and Borders, supra note 1.
8
9
Of course, one is always in real-space, so this may seem to be a false
dichotomy. The key is that there is a difference between acting in cyberspace
while living in real-space and acting exclusively in real-space.
10
“Second Life” is a virtual world with 100,000 residents that allows users to
create their own content. See Second Life: Your World, Your Imagination,
http://secondlife.com/ (last accessed
11
The blogosphere is comprised of all weblogs on the internet. It is a kind of
self-defining “place” in that it is the place where all blogs are located, but
it is no more than the sum of those blogs. See Blogosphere – Wikipedia,
http://en.wikipedia.org/wiki/Blogosphere (last accessed
12
It may be possible to trace IP addresses, server locations, and the like to get
some idea of where cyberspace activities take place. But the massive scale of
cyberspace and the number of servers, sites, transactions, and users render
useful, consistent knowledge of real-space locations impractical, if not
impossible.
13
Johnson & Post refer to this as “a new ‘mind-body’ problem on a global
scale.” Johnson & Post, Law and Borders, supra note 1, at
1378.
14
15
16
17
This comes from David G. Post on “The Great Debate” Panel at the 2005 State of
18
Johnson & Post, Law and Borders, supra note 1, at 1380.
19
Johnson & Post are not alone in making this claim. See, e.g., Joel
R. Reidenberg, Governing Networks and Rule-Making in Cyberspace, in BORDERS
IN CYBERSPACE: INFORMATION POLICY AND THE GLOBAL INFORMATION INFRASTRUCTURE 84,
100 (Brian Kahin & Charles Nesson eds., 1997).
20
There is some question as to whether these are possible. What matters, for my
purposes, is the architecture of cyberspace is not necessarily fixed. See
21
22
Johnson & Post, Law and Borders, supra note 1, at 1387-91
23
Jonathan Zittrain, for one, believes that we could see a much less democratic
internet in the future. See, e.g., Jonathan Zittrain, Be Careful What
You Ask For: Reconciling a Global Internet and Local
Law, in WHO RULES THE NET? INTERNET GOVERNANCE AND JURISDICTION AND
JURISDICTION 13 (Adam Thierer & Clyde Wayne Crews Jr. eds., 2003) at 29
(arguing, “What we might gain in easing jurisdictional tensions we could stand
to lose in revolutionary capacity”).
24
Goldsmith, supra note 1. See also Bruce Keller, The Game's the
Same: Why Gambling in Cyberspace Violates Federal Law, 108 YALE L.J. 1569,
1572 (1999); Lessig supra note 20 (arguing that a complete separation
between real-space law and cyberlaw is impossible); Timothy Wu, When Law
& the Internet First Met, 3 GREEN BAG 2D 171 (2000) (arguing that
Johnson & Post’s vision has never come to fruition). But see Post, Against
“Against Cyberanarchy”, supra note 1.
25
See, e.g., Goldsmith, supra note 1, at 1235 (“[T]he choice-of-law
issues implicated by cyberspace transactions are no more complex than the
issues raised by functionally identical multijurisdictional transactions that
occur in real space all the time.”).
26
Frank H. Easterbrook, Cyberspace and the Law of the Horse, 1996 U. CHI.
LEGAL. F. 207 (1996). But see Lawrence Lessig, The Law of the Horse:
What Cyberlaw Might Teach, 113 HARV. L. REV. 501
(1999).
27
Easterbrook, supra note 26, at 207.
28
Post, Against “Against Cyberanarchy”, supra note 1, 1373-76.
29
30
Goldsmith, supra note 1, at 1200.
31
Perhaps this is why Post, even in his response to Goldsmith, does not mention
this distinction. See Post, Against “Against Cyberanarchy”, supra
note 1.
32
33
Post, for example, takes issue with Goldsmith’s term “regulation skeptics.” See
Post, Against “Against Cyberanarchy”, supra
note 1, at 1368.
34 JOHN LOCKE, THE SECOND TREATISE OF GOVERNMENT, in
TWO TREATISES OF GOVERNMENT (Peter Laslett ed., Cambridge Univ. Press
1988), § 95 (emphasis in original).
35
36
37
For a discussion of the relationship between the natural law and civil law, see
Steven J. Horowitz, Rethinking Lockean Copyright and Fair Use, 10
DEAKIN L. REV. 209, 213 (2005).
38
LOCKE, supra note 34, § 14.
39
There may be contractual terms that obviate some problems, and treaties between
countries to solve others. What matters is not how this situation is resolved
but that it exists.
40
LOCKE, supra note 34, § 4.
41
42
For a discussion of the natural common, see Horowitz, supra note
37, at 214.
43
Refuting Robert Filmer’s justification for absolute power was one of Locke’s
primary motivations in writing the Treatises. See ROBERT FILMER, Patriarcha,
in PATRIARCHA AND OTHER POLITICAL WORKS OF SIR ROBERT FILMER 47-126 (Peter
Laslett ed., Cambridge Univ. Press 1949).
44
See LOCKE, supra note 34, § 19-21.
45
Rousseau, for one, suggests that it was not necessary to form the first
societies. See JEAN-JACQUES ROUSSEAU, DISCOURSE ON THE ORIGIN OF
INEQUALITY (Patrick Coleman ed., 1994).
46
What this consent consists in, and whether there can be tacit consent,
is a general problem for social contract theory that is beyond the scope of
this paper. For a discussion of the problems of tacit consent for social
contract theory, see Peter J. Steinberger, Political Obligations and
Derivative Duties, 64 J. POL. 449 (2002).
47
An analysis of the political state of cyberspace appears to presuppose the
contentious claim that cyberspace is a place. While I think it makes sense to
think of cyberspace as a place, I will not need to treat it as such for my
analysis. Because the state of nature describes a type of relationship between
people (not places), it does not matter whether people in cyberspace are really
in cyberspace at all. That said, I should note that the analysis does
become cumbersome without relying on place metaphors, so I will rely on words
that connote a sense of place. I do this to avoid overcomplicating the issues
at hand, but I am not committed to their underlying assumptions. Scholars have
devoted substantial effort to the “cyberspace as place” question. See, e.g.,
Clay Calvert, Regulating Cyberspace: Metaphor, Rhetoric, Reality, and
the Framing of Legal Options, 20 HASTINGS COMM. & ENT. L.J. 541 (1998); Dan Hunter, Cyberspace as Place and the Tragedy
of the Digital Anticommons, 91 CALIF. L. REV. 439 (2003); Keller, supra
note 24; Mark A. Lemley, Place and Cyberspace, 91 CALIF. L. REV. 521
(2003); Lessig, supra note 20.
48
LOCKE, supra note 34, § 14.
49
For discussions of law in virtual worlds, see F. Gregory Lastowka &
Dan Hunter, The Laws of the Virtual Worlds, 92 CALIF. L. REV. 1 (2004); Institute
for Information Law and Policy Symposium: State of
50
Johnson & Post argue, for example, that citizens of cyberspace will make
their own rules. See Johnson & Post, Law and Borders, supra
note 1, at 1387-91.
51
Reidenberg has called this a “semi-sovereign status.” See Reidenberg, supra note 19, at 100.
52
Thanks to Nathaniel Fleischer for suggesting this line of criticism.
53
For a proper Lockean framework, we are missing an impartial judge and a much
more complicated government. For simplicity, I have left these out, but of
course I could design a world with these as well, still with no appeal to
real-space sovereignty.
54
You have to go to the meetings.
55 Supra, Part II.B.
56
Frontier metaphors are common. See, e.g., Capitanchik & Whine, supra
note 6 at 253 (describing the internet as “the last bastion of the absolute
right to free speech”); Thomas W. Evans & Howard G. Zaharoff, Cyberspace
and the Internet: Law’s Newest Frontier, 41 B. B. J. 14 (1997); Llewellyn
Joseph Gibbons, No Regulation, Government Regulation, or Self-Regulation:
Social Enforcement or Social Contracting for Governance in Cyberspace, 6
CORNELL J.L. & PUB. POL'Y 475 (1997); Raymond Ku, Foreword: A Brave New
Cyberworld?, 22 T.
57
See Gene J. Koprowski, China Internet Use Catching Up
with US, TECHNEWSWORLD (
58
Locke would say that Natural Law commands universally so that there are some
laws that everyone must obey. See, LOCKE, supra note 34, § 135; see
also Horowitz, supra note 37, at 213.
59
60
Michael Geist has a three-part test to determine jurisdiction, the first of
which—the existence of a contract between two parties—is similar to what I
present here. See Michael Geist, The Shift Toward “Targeting” for
Internet Jurisdiction, in WHO RULES THE NET? INTERNET GOVERNANCE AND
JURISDICTION supra note 23, at 91, 108 (“The first [factor]…considers
whether either party has used a contractual arrangement to specify which law
should govern.”).
61
Supra Part II.
62
While I argue that real-space geography has little to do with cyberspace
interactions, Monroe Price has argued that geography and media inevitably
influence each other. See
63
Barlow, supra note 3.
64
David Post provides an excellent argument for the importance of scale, citing
the Netcom case in which the
65
According to Internet World Stats, there were 972,828,001 internet users as of
66
“Force” could mean any number of things, from real-space political action
(e.g., embargos) to cyberspace action (e.g., Denial of Service Attacks).
67
LOCKE, supra note 34, § 19 (emphasis in original).
68
LOCKE, supra note 34, § 125 (“[T]hough in the
state of Nature he hath such a right, yet the Enjoyment of it is very
uncertain, and constantly exposed to the Invasion of others.”).
69
See supra IV.B.
70
Barlow, supra note 3 (saying to real-space governments: “I declare the
global social space we are building to be naturally independent of the
tyrannies you seek to impose on us. You have no moral right to rule us.”).
71 Supra, Part IV.