10/9/02
Question 1
Regulatory Workarounds of Open Source Software on the
Net
In the Bible, Adam and Eve lived in the Garden of Eden and experienced a life free from every care in the world -- free from sin, free from hunger, free from disease. Lurking in this Garden of Eden, however, was a serpent – “more subtle than any best of the field” (Genesis 3:1) who tricks Eve into eating the apple from the Forbidden Tree. Eve in turn offers the fruit to Adam. Both have now eaten from the Forbidden Tree and have now committed the Original Sin. Consequently, God casts Adam and Eve out of the Garden of Eden never to return to this Utopian land.
Lawrence Lessig, in Code, describes the
inevitable and “subtle beast” of the Market (and his friend Government) as the
means for more and more regulation of the Net.
He then proposes that open source software will work to stifle this
regulation. One would be incorrect,
however, to deduce that a Net compromised entirely of open source
software will make the Net unregulable.
Open source would merely place a “check” or “balance” on the regulation,
but would not completely stifle it.
Open source software greatly deflates the power or direct
regulation, but indirect regulation, can still occur—and to a great
extent.
First of all, we need to understand what factors play
into regulating the actions performed on the Net, and which factor we will
primarily be concerned with in this paper.
Lessig proposes four constraints: social norms, law (government), the
market, and the architecture of the space.[1] To put it briefly, social norms consist of
the behaviors of which society approves and disapproves. Law consists of all the laws government has
enacted that affect a citizen’s actions.
The market can place price constraints or advertise in popular (or not
advertise in unpopular) domains to influence the user. The architecture of a place details what is
possible/impossible in that place. In a
civilized state the government and market play a large rule in regulating the
actions of the citizens. In this paper,
we will focus on how the market, combined with help from the government, tries
to increase its control over the Net, and then determine the effect open source
software has on this control.
In order for the Net to be effectively regulated,
Lessig asserts that the Net needs to possess two qualities. First that there needs to be a small number
of targets for regulation. If there is
a small number of these targets, then the government can efficiently oversee
that these targets are complying with its wishes. Secondly, the targets need to possess control over the users of
its network. If the points of contact
alter a feature of the Net, then those who use the Net will unavoidably be
affected by this change.
To what extent does open source have on these two
conditions? Proponents of open source
maintain that having the code of the Net be open source eliminates these two
conditions, thus rendering the Net unregulable. To a certain extent, this is true. Having code be open source greatly multiplies the number of
targets of regulation. The government
cannot efficiently monitor this large number of targets. Thus, open source software debunks the first
requirement for regulation. Open source
software also strikes down the second condition for regulation: “useful
targets.” Since the code is free to all
to look at and modify, there is not one source of control; the one-to-many
relationship no longer exists. The
government can mandate a certain feature in a program, but it’s up to the
millions of users to decide whether or not they want to update their version
for this new feature (or they can modify the feature). Open source code thus eliminates the two
necessary conditions for regulation.
So open source makes the Net unregulable, right?
Wrong. It only makes direct
regulation ineffective, but the market/government can still use indirect
means for effective regulation. Digital
certificates, through cryptography, can enable the government indirectly regain
the two necessary conditions for regulation: small number of targets and useful
targets. Via digital certificates, a
person can be acceptably confident that the person he is talking with is who
that person says he is. This trust in
authentication is granted from encryption.
This is an essential requirement for e-commerce; if a marketplace can be
reasonably confident that their customer is legit, it can go forth with its
business. Government can take advantage
of this need for digital certificates to enact its own regulations on the
Net. The government can require that
important/widely used commercial services comply with its wishes. It can make a bank, for example, only
interact with users whose certificate contains certain data – criminal record,
birth record, citizenship, etc. – the requirements can potentially have nothing
to do with a bank transaction. In other
words, the government regulates the intermediaries (the banks), not the
millions of people (the users of the banks).
The number of intermediaries is far fewer, and their interests are
economical and not typically political, so they will most likely comply with
the government’s wishes. In effect, the
government has reasserted condition number one: a small number of targets.
Now onto regaining condition number two: targets that
have a one-to-many control. By choosing
to regulate intermediaries that have an important impact on many individual’s
lives, the government can effectively enact its wish of regulation on many
individuals. If everyone who wants to
interact with a bank, or visit a governmental website, or buy books online must
have a ID with specific information, then life without such an ID will be very
cumbersome. The user is “forced” into
“voluntarily” accepting the ID in order to perform basic functions on the
Net. In essence, “the incentive to
carry ID is so strong that no government requirement is necessary” (Code
50). Thus, even though the code of the
Net is open source and modifiable, users may still be “required” to use
features and protocols specified by the government. In this sense, the government has to power to regulate the
Net.
In addition to digital certificates with the aid of cryptography, changing the location where code resides can lessen open source’ ability to hinder regulation. If crucial segments of code were taken out of the software realm and incorporated in the hardware of a computer, then open source’s ability to debunk the one-to-many requirement of regulation is itself eliminated. Palladium is such an example. The code will still be open source, but even the savvy computer scientist will not likely be able to modify the hardware code. And even if the scientist could, he would probably perform the laborious task on just his own machine. Combining cryptography with hardware code can potentially render the “open source” code unmodifiable, thus taking regulation out of the hand of the individual.
The advantages of open source software, however, are
not all lost. Hindering direct
regulation offers a valuable resource to the Net community; it makes the
actions and the actors of direct regulation visible. The government and market can no longer “hide” under closed code
to promote the features they want to see.
If the government wants to alter the performance of an application, then
having it open source shows to the world what exactly the government wants to
do and also that the government is the one who wants to do it. In this sense, open source offers a “check”
on government regulation. Although this
“check” cannot overrule decisions, but it takes away the luxury that the
government can act under the cover of closed source.
Open source software provides a hindrance to direct
regulation of the Net, but it does not effectively disable indirect regulation.
With the aid of cryptography and
hardware implementation of code, indirect regulation can be highly
effective. But open source still does
have its redeeming qualities. It makes
the regulators transparent and their actions transparent. Open source does not make the Net entirely
unregulable, but it makes regulation more difficult and less effective.
[1] (“Space” here will refer to the Internet but it is more generally applied to the entity where the action takes place. For example, if you want to prevent people from reaching the roof of a building, you can demolish the stairs the lead to the roof—thus changing the building’s architecture.)