ssrn id310020

THE GEORGE WASHINGTON UNIVERSITYLAW SCHOOL

PUBLIC LAW AND LEGAL THEORY WORKING PAPER NO. 382002

The Problem of Perspective in Internet Law

Professor Orin S. Kerr

Georgetown Law Journal, Vol. 91, February 2003

This paper can be downloaded without charge from theSocial Science Research Network Electronic Paper Collection:

http://ssrn.com/abstract_id=310020

http://ssrn.com/abstract_id=310020

The Problem of Perspective in Internet Law

ORIN S. KERR*

INTRODUCTION

The lawyer’s quintessential task is to apply legal rules to facts.1 When weapply law to the Internet, however, a difficult question arises: What are the“facts” of the Internet? The Internet’s facts depend on whether we look tophysical reality or virtual reality for guidance. We can model the Internet’s factsbased on virtual reality, looking from the perspective of an Internet user whoperceives the virtual world of cyberspace and analogizes Internet transactions totheir equivalent in the physical world.2 Alternatively, we can model the factsbased on the physical reality of how the network operates. From this perspec-tive, Internet transactions can be understood based on how the network actuallyworks “behind the scenes,”3 regardless of the perceptions of a user. Because theInternet can generate a virtual reality, it offers two distinct sets of facts: onebased on physical reality, the other based on virtual reality.

The Internet’s ability to generate a virtual reality creates what I will call theproblem of perspective in Internet law. The problem is that whenever we applylaw to the Internet, we must first decide whether to apply the law to the facts asseen from the viewpoint of physical reality or virtual reality. In this Article, Iwill refer to the viewpoint of virtual reality as the “internal perspective” of theInternet, and the viewpoint of physical reality as the “external perspective.”

This Article argues that the problem of perspective pervades Internet law, andthat the nature and shape of Internet law depends upon how it is resolved inparticular cases. In a surprising number of situations, we arrive at one resultwhen applying law from an internal perspective and a different result whenapplying law from an external perspective. In fact, many of the major disputeswithin the field of “cyberlaw”4 boil down to clashes between internal andexternal perspectives. To complicate matters, neither perspective holds an apriori claim to greater legitimacy. Both internal and external perspectives canappear perfectly viable depending on the circumstances, and courts and commen-

* Associate Professor, George Washington University Law School. Thanks to Rachel Barkow, AmyBarrett, Laura Berger, James Boyle, Robert Brauneis, Brad Clark, Erica Hashimoto, Dan Hunter, MarkLemley, Doug Lichtman, Chip Lupu, Dan Markel, Jonathan Molot, Richard Pierce, Jeffrey Rosen,Michael Selmi, Joe Singer, Daniel Solove, Peter Swire, Molly Van Houweling, and Tim Wu, as well asthe participants in the George Washington University Law School faculty workshop. I benefited greatlyfrom presenting early versions of this piece before the law faculties at George Washington, Northwest-ern, Rutgers-Newark, Cardozo, Washington University, and Boston College.

1. See, e.g., EDWARD H. LEVI, AN INTRODUCTION TO LEGAL REASONING 1–3 (1949).2. See infra notes 8–9 and accompanying text.3. See infra notes 10–12 and accompanying text.4. See Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113 HARV. L. REV.

501, 501 (1999).

357

tators switch between them frequently without even recognizing the change.5

The goal of this Article is to explain the problem of perspective, to show itsimportance, and to offer an approach that can help lead to its solution. I arguethat we need to be aware of the problem of perspective and develop legal toolsthat can help us choose between real and virtual understandings of the Internetwhen we apply law to it. I also contend that the problem of perspective is a newproblem, or at least a new twist on an older one. Although prior technologiessuch as the telephone hinted at the clash in narrow ways,6 computer networksand the Internet provide the first widespread technology that creates a virtualworld for its users that can compete on an equal footing with the real one. As aresult, Internet law prompts us to confront the problem of perspective for thefirst time.

I will develop my argument in four parts. In Part I, I explain the conceptualdistinction between internal and external understandings of the Internet. ThePart explores how and why both perspectives offer plausible means of modelingthe facts of the Internet and how the shape of Internet law hinges on whether weapproach the law of the Internet internally or externally.

Part II shows that the choice between internal and external perspectivespresents a recurring and powerful theme in Internet law. This Part explores theimportance of the problem of perspective in four areas of law. The analysisbegins with the Fourth Amendment in cyberspace, and in particular FourthAmendment restrictions on obtaining e-mail and conducting remote networksearches; turns next to Internet governance, with particular focus on the ideas ofLawrence Lessig and the scope of the state action doctrine; follows withcomputer crime law, where it shows how internal and external approaches havebeen raised in appellate cases involving online bomb threats and the distributionof obscene materials over the Internet; and concludes with copyright law,focusing on the recent copyright dispute over the Internet music serviceMP3.com. Within each area, I show how a range of seemingly unrelateddisputes in fact all reflect a clash between internal and external viewpoints.

Part III explains the significance of the problem of perspective. It shows howthe problem of perspective reflects the universalization of themes presaged inlimited ways by the telephone and builds upon and extends theories of systemsdeveloped by H.L.A. Hart and others. This Part also argues that we need todevelop a new awareness of the problem of perspective, as well as a frameworkto help mediate between the internal and external perspectives in particularcases.

Part IV offers a tentative normative framework for choosing which perspec-tive to apply in a given case. The framework approaches the problem ofperspective as a problem of fidelity, in which our goal should be to select the

5. See infra Part II.6. I discuss the precursors to the problem of perspective in prior technologies such as the telephone

and the telegraph in section III.A.

358 [Vol. 91:357THE GEORGETOWN LAW JOURNAL

perspective that best reflects the judgments and function of preexisting law.Under this framework, our first step should be to study the legal doctrine forsignals that the law already reflects an internal or external approach. Whenapplicable cases or statutes reveal a particular orientation, we should adopt thatorientation. When doctrine alone does not answer which perspective shouldapply, our second step should be to adopt the perspective of the party that thelaw seeks to regulate. As with the first method, this approach is rooted in fidelityto existing law: It seeks to transfer the law’s regulatory scheme as directly andcompletely as possible from the physical world to the Internet.

I. THE INTERNAL AND EXTERNAL PERSPECTIVES IN INTERNET LAW

A. THE PROBLEM OF PERSPECTIVE

In the 1999 science fiction thriller The Matrix,7 Keanu Reeves plays acomputer hacker named “Neo” who learns that the reality he has known sincebirth is merely a virtual reality created by a computer network known as theMatrix. The real Neo lies in a semicomatose state attached to the network, towhich he and others have been connected by advanced computers that havetaken over the world and sap energy from humans while occupying their mindswith virtual reality. Neo ends up joining the rebel forces trying to destroy theMatrix, and the movie jumps several times between the virtual world inside theMatrix and the real world outside of the Matrix. The movie presents us with twodifferent realities, two existing worlds. The first reality is the virtual world thatwe experience inside the Matrix, and the second is the “real” world that weexperience outside the Matrix.

In addition to being a fun movie, The Matrix points out an important problemthat arises when we try to understand the nature of computer networks ingeneral and the Internet in particular. Like Neo confronting the Matrix, we canthink about the Internet in two ways, virtual and real. The virtual perspective islike the perspective inside the Matrix: it accepts the virtual world of cyberspaceas akin to a reality. Of course, unlike Neo, we know all along that the virtualworld that the computer generates is only virtual. But as we try to make sense ofwhat the Internet is, to understand what we experience online, we might decideto treat that virtual world as if it were real.

I will call this virtual point of view the internal perspective of the Internet.The internal perspective adopts the point of view of a user who is logged on tothe Internet and chooses to accept the virtual world of cyberspace as a legiti-mate construct.8 To this user, a computer connected to the Internet provides awindow to a virtual world that is roughly analogous to the physical world of

7. THE MATRIX (Warner Bros. Pictures 1999). Warner Bros. maintains a website for the movie athttp://whatisthematrix.warnerbros.com (last visited Jan. 5, 2003).

8. John Perry Barlow helped popularize the term “cyberspace” with his “Declaration of the Indepen-dence of Cyberspace.” See http://www.eff.org/�barlow/Declaration-Final.html (last visited Jan. 5,2003) (“Governments of the Industrial World, you weary giants of flesh and steel, I come from

2003] 359THE PROBLEM OF PERSPECTIVE IN INTERNET LAW

real space. The user can use her keyboard and mouse to go shopping, send mail,visit a chat room, participate in an online community, or do anything else shecan find online.9 The technical details of what the computers attached to theInternet actually do “behind the scenes” don’t particularly matter. What mattersis the virtual world of cyberspace that the user encounters and interacts withwhen he or she goes online.

We can also understand the Internet from a different perspective. Like Neowhen he is outside the Matrix, we can look at the Internet from the point ofview of the physical world, rather than the virtual one. I will call this theexternal perspective of the Internet. The external perspective adopts the view-point of an outsider concerned with the functioning of the network in thephysical world rather than the perceptions of a user.

From this external viewpoint, the Internet is simply a network of computerslocated around the world and connected by wires and cables.10 The hardwaresends, stores, and receives communications using a series of common proto-cols.11 Keyboards provide sources of input to the network, and monitors providedestinations for output. When the Internet runs properly, trillions of zeros andones zip around the world, sending and receiving communications that thecomputers connected to the network can translate into commands, text, sound,and pictures.

From the external perspective, the fact that Internet users may perceive thatthey have entered a virtual world of cyberspace has no particular relevance.These perceptions reflect the fact that software designers often garnish theirapplications with icons, labels, and graphics to help novices understand and usethem—for example, by writing e-mail programs so that e-mail looks and feelslike postal mail.12 These superficialities have no deeper meaning from theexternal perspective. What matters is the physical network and the technicaldetails of how it works, not the easily manipulated perceptions of Internet users.

Both internal and external understandings of the Internet should ring true tomost of us. The Internet is a physical network, and it can create a virtual worldfor its users that can appear sufficiently realistic to its users to make a plausibleclaim for equal footing with the physical world.13 But the key for us is that by

Cyberspace, the new home of Mind. On behalf of the future, I ask you of the past to leave us alone. Youare not welcome among us. You have no sovereignty where we gather.”).

9. See David G. Post, “The Free Use of Our Faculties”: Thomas Jefferson, Cyberspace, and theLanguage of Social Life, 49 DRAKE L. REV. 407, 410 (2001).

10. See Reno v. Am. Civil Liberties Union, 521 U.S. 844, 849 (1997) (describing the Internet as “aninternational network of interconnected computers.”).

11. See PRESTON GRALLA, HOW THE INTERNET WORKS 6–7 (1999).12. See NATHAN J. MULLER, DESKTOP ENCYCLOPEDIA OF THE INTERNET 87–95 (1999).13. See Charles Nesson & David Marglin, The Day the Internet Met the First Amendment: Time and

the Communications Decency Act, 10 HARV. J. L. & TECH. 113, 121 (1996) (“The Internet at present ismostly text and still photo, but again, the situation is changing daily. Its future promises the super-charged impact of virtual reality and interactivity.”); David R. Johnson & David Post, Law andBorders—The Rise of Cyberspace, 48 STAN. L. REV. 1367, 1367 (1996) (“Cyberspace challenges the


generating a virtual reality, the technology in a sense leaves us with twoInternets, rather than one.14 We have an external version of the Internet, andalso an internal one. One is physical, the other virtual.15

B. PERSPECTIVE AS A PROBLEM OF LAW

Why does this matter to lawyers and to the nature of Internet law? It mattersbecause legal outcomes depend on facts, and the facts of the Internet depend onwhich perspective we choose.16 This is a very practical problem. The basic taskof a lawyer is to apply legal rules to facts—to apply law to an understanding ofreality. In the case of the Internet, however, two competing understandings ofreality exist. We have a virtual reality from the internal perspective and aphysical reality from the external perspective. This means that we face a choiceof which perspective to use when applying law to the Internet. Do we decide tofollow the internal perspective of virtual reality or the external perspective ofphysical reality? Which version of the Internet should we pick before applyingthe law to it? By choosing the perspective, we choose the reality; by choosingthe reality, we choose the facts; and by choosing the facts, we choose the law.

We can look at this another way by noting the differences between whathappens when we apply law to the Internet from an internal versus an externalperspective. From the internal perspective of an Internet user, the Internet iscyberspace, and we apply law to the Internet by trying to map the physicalworld of “realspace” onto the virtual world of cyberspace.17 We look foranalogies between cyberspace and realspace, and try to match the rules betweenthem.18 To the external observer, in contrast, the Internet is the physicalnetwork, and we apply law to the Internet by applying the law to the electronictransactions underlying the network’s operation. This does not necessarily meanthat the Internet must be viewed only as 0s and 1s, any more than modeling the

law’s traditional reliance on territorial borders; it is a ‘space’ bounded by screens and passwords ratherthan physical markers.”).

14. In an earlier essay, I attempted to describe a similar phenomenon by referring to Georges Seurat,whose paintings of nineteenth century Paris consist entirely of small dots. See Orin S. Kerr, Are WeOverprotecting Code? Thoughts on First-Generation Internet Law, 57 WASH. & LEE. L. REV. 1287,1287 (2000). We can understand Seurat’s paintings in two ways: We can look at them “close-up” as justa collection of dots, or else we can step back and see them as depictions of nineteenth century Paris. Iargued that the Internet is similar: we can see it as just bits and bytes, like the close-up perspective ofSeurat’s paintings, or else we can see it as the virtual world of cyberspace, like viewing a Seuratpainting from a distance. See id. at 1287–88. The correspondence between my former and currentapproach is not exact. However, the “close-up” version of the Internet roughly reflects an externalperspective, and the version at a distance generally tracks the internal perspective.

15. Cf. Timothy Wu, Essay, Application-Centered Internet Analysis, 85 VA. L. REV. 1163, 1189 n.70(1990) (noting the distinction between application layers and networks layers as represented in theInternational Organization for Standardization’s Open Systems Interconnection model).

16. See, e.g., JEROME FRANK, LAW AND THE MODERN MIND xii–xiv (1949).17. See Dan Hunter, Cyberspace as Place, and the Tragedy of the Digital Anticommons, 91 CAL. L.

REV. (forthcoming February 2003) (discussing applications of Internet law that view cyberspace “as aplace.”); Mark Lemley, Place and Cyberspace, 91 CAL. L. REV. (forthcoming February 2003) (same).

18. See Kerr, supra note 14, at 1289.


physical world requires us to model sounds as pressure waves or light asphotons of energy. But it does mean that we look for analogies betweenrealspace and the behind-the-scenes action that the computers connected to theInternet process and complete.

These two approaches are similar to each other and also quite different. Inboth the external and internal cases, we apply law to “the Internet.” However,our model of what that Internet is—and therefore what Internet law is—variesdramatically depending on the perspective we choose. The law is contingent onthe facts, and the facts are contingent on our perspective.

What makes this problem unusually interesting is that there is no particularcorrelation between internal and external renderings of the Internet’s facts. Thereal produces the virtual, but the virtual need not reflect the real. Significantchanges in the behind-the-scenes workings of the Internet can go entirelyunnoticed by users.19 At the same time, minor changes in computer code canhave a dramatic impact on users’ experiences.20 A typical user immersed in theinternal perspective can be blissfully unaware of the complex inner working ofthe Internet.

The lack of correlation between the real and the virtual has profound implica-tions for Internet law. It means that the legal outcomes reached using an internalset of facts exist independently from outcomes reached with an external set offacts. When we apply the law to the facts, an internal perspective will take usdown one path, and an external perspective will take us down another. The twopaths may happen to converge, but there is no reason to think they will. Ineffect, we not only have two Internets, but two versions of Internet law. Everytime we apply law to the Internet, we will have two possible outcomes: aninternal outcome and an external outcome. The two outcomes may happen tomatch in some cases. In many cases, however, the choice of perspective provesoutcome-determinative. Consequently, the shape of Internet law hinges on ourchoice of perspective.

C. AN EXAMPLE: SURFING THE WEB

All of this may seem rather abstract, so an example may help. Consider whathappens when an Internet user surfs the web. Imagine that an Internet useropens up a web browser and types in “www.amazon.com,” and moments laterthe homepage of Amazon.com appears on the viewer’s screen. How might wemodel this event? How can we develop a factual picture of what has happened,so that we can later determine the legal consequences of accessing a webpage?

This is easy from an internal perspective. The user has visited Amazon.com’s

19. The differing operations of popular e-mail programs provide an apt comparison. E-mail pro-grams can work in fairly different ways, but those contrasts are unknown to the casual user. SeeMULLER, supra note 12, at 91–95.

20. See LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE 6 (1999) (explaining that “thesoftware and hardware that make cyberspace what it is regulate cyberspace as it is”).


website, going to Amazon.com’s home on the Internet. The user has visitedAmazon.com’s virtual store much like a person might visit a store in thephysical world, traveling from one point in cyberspace to another. Of course, werealize that the user has not actually traveled anywhere. The user is just sittingin front of the screen. But from an internal perspective, the essential experienceof surfing Amazon.com can be captured by comparing it to visiting a store.

From an external perspective, however, the event appears quite different—and significantly more complicated. Behind the scenes, the simple act of typing“www.amazon.com” into a web browser triggers a series of responses fromdifferent computers connected to the Internet. The browser begins by sendingout a request across the Internet to a special type of computer known as aDomain Name System (DNS) server.21 The browser’s request asks the DNSserver to translate the letters of the website address “amazon.com” into an“Internet Protocol” or “IP” address, which is a series of numbers that computersconnected to the Internet understand as an address akin to a phone number.22

The DNS server will respond that “www.amazon.com” translates into the IPaddress “207.171.184.16.”23 The user’s browser then issues another request, thistime directed to “207.171.184.16,” asking it to send a set of data files back tothe browser. Amazon.com’s computer will receive the request and then senddata back to the browser. The browser will receive the data and display it on theuser’s screen. The resulting images and text appear in the form of the Amazon.com webpage that the user requested.24

Notice that the internal and external perspectives have produced two differentaccounts of the same event. One model of the facts follows the virtual perspec-tive of the user, and another model follows the behind-the-scenes perspective ofhow the Internet actually works. From the internal perspective, visiting Amazon-.com resembles visiting a store. The user types in the address, and a momentlater is paying a virtual visit to Amazon.com’s site. From the external perspec-tive, visiting Amazon.com resembles calling Information and asking for Amazon.com’s phone number, then dialing the number and asking the representative tosend you the latest Amazon.com catalog. The single event of surfing the webproduces two set of facts, one internal and the other external. As a result, whenwe need to apply law to the act of visiting a website, we can apply that law totwo different sets of facts, which can produce two different outcomes.25

Note that I have not yet mentioned legal theory. Much of the scholarship on

21. See MULLER, supra note 12, at 71–74 (explaining the function and configuration of DNSservers).

22. See id. at 222–23 (explaining IP addresses).23. This can be verified in several ways, among them by typing in “207.171.184.16” into any web

browser that is connected to the Internet. The browser will direct the user to “http://www.amazon.com.”24. See MULLER, supra note 12, at 222–23.25. See Maureen O’Rourke, Property, Rights and Competition on the Internet: In Search of an

Appropriate Analogy, 16 BERKELEY TECH. L.J. 561, 580 (2001) (“[C]ourts emphasize different factorsdepending on whether they focus on the website . . . or the tangible server on which it resides . . . .”).


Internet law focuses on normative questions of theory: on what the law shouldbe and how it should or should not change when it moves from the physicalworld to the Internet.26 These are vitally important questions. However, theproblem of perspective arises before we reach them. Before applying the law,we must first develop a mental picture of existing facts upon which the lawcould apply. The problem of perspective focuses on how we develop that mentalpicture. Although theory may be needed to resolve the problem of perspective,we do not need theory to recognize it.

II. THE POWER OF PERSPECTIVE: FOUR EXAMPLES

This Part reveals the importance of perspective by showing how it influencesthe shape and nature of Internet law across a wide range of substantive areas.The analysis showcases four unrelated fields of law and explains how theproblem of perspective affects the shape of law within each. It begins with theFourth Amendment in cyberspace, and in particular Fourth Amendment restric-tions on obtaining e-mail and conducting remote network searches; turns next toInternet governance, with particular focus on the ideas of Lawrence Lessig andthe scope of the state action doctrine; follows with computer crime law, where itshows how internal and external approaches have been raised in appellate casesinvolving online bomb threats and the distribution of obscene materials over theInternet, as well as in the Supreme Court’s decision in Reno v. American CivilLiberties Union;27 and concludes with copyright law, focusing on the recentcopyright dispute over the Internet music service MP3.com.

The examples in this Part do not present an exhaustive list of legal disputestriggered by the clash between internal and external perspectives. To thecontrary, the examples merely offer a few glimpses at a broader phenomenonthat occurs when law meets the Internet. By studying these examples, however,we can appreciate the importance of the problem of perspective and becomeaware of how it influences Internet law.

A. THE FOURTH AMENDMENT IN CYBERSPACE

The Fourth Amendment law of search and seizure offers a rich panoply ofsituations in which the distinction between internal and external perspectivestakes on critical importance.28 More than most areas of law, Fourth Amendmentjurisprudence is heavily spatial:29 the rules speak of rights to access somespaces but not others,30 and the constitutionality of a search often hinges on the

26. See sources cited infra notes 129–33.27. 521 U.S. 844 (1997).28. See U.S. CONST. amend. IV.29. See Erik G. Luna, Sovereignty and Suspicion, 48 DUKE L. J. 787, 838 (1999) (noting the “spatial

orientation” of the Fourth Amendment).30. For example, search warrants can be issued only to search specific places for specific evidence.

See Walter v. United States, 477 U.S. 649, 657 (1980) (noting that “[b]ecause indiscriminate searchesand seizures conducted under the authority of ‘general warrants’ were the immediate evils that


type of space involved.31 As a result, the choice between internal and externalperspectives presents a recurring theme of the Fourth Amendment in cyber-space.

I will explore the impact of perspective on the Fourth Amendment in cyber-space by examining two hypotheticals. Both raise important questions that thecourts are just beginning to confront and have not yet resolved. The firsthypothetical considers whether the Fourth Amendment requires the police toobtain a search warrant to obtain e-mail from an Internet service provider. Thesecond hypothetical asks whether a search warrant that authorizes the search ofa computer connected to a network also implicitly authorizes the search ofremotely stored files that are virtually present on the network.

1. Do the Police Need a Warrant to Obtain E-mail?

Imagine that A sends an e-mail to his friend B. Two police officers learnabout the e-mail and believe that it may reveal a nefarious criminal conspiracy.The officers agree that they should try to obtain a copy of the e-mail to provethe conspiracy. They confront a legal question: what kind of legal process mustthey follow to obtain the e-mail? Does the Fourth Amendment require them toobtain a search warrant? Or can they obtain the e-mail with less process than asearch warrant? The answer depends largely upon whether they apply aninternal or external perspective of the Internet.

Imagine that the first officer applies an internal perspective of the Internet. Tohim, e-mail is the cyberspace equivalent of old-fashioned postal mail. Hiscomputer announces, “You’ve got mail!” when an e-mail message arrives andshows him a closed envelope.32 When he clicks on the envelope, it opens,revealing the message. From his internal perspective, the officer is likely toconclude that the Fourth Amendment places the same restriction on governmentaccess to e-mail that it places on government access to ordinary postal mail. Hewill then look in a Fourth Amendment treatise for the black letter rule onaccessing postal mail. That treatise will tell him that accessing a suspect’s mailordinarily violates the suspect’s “reasonable expectation of privacy,” and thattherefore the officer must first obtain a warrant.33 Because e-mail is the equiva-

motivated the framing and adoption of the Fourth Amendment . . . that Amendment requires that thescope of every authorized search be particularly described.”) (internal quotations and citations omitted).

31. See generally Ronald J. Allen & Ross M. Rosenberg, The Fourth Amendment and the Limits ofTheory: Local Versus General Theoretical Knowledge, 72 ST. JOHN’S L. REV. 1149 (1998) (noting thatthe Fourth Amendment case law has created very specific and distinct rules for different spaces).

32. A sound recording of a voice saying “You’ve got mail!” is played whenever an America Online(AOL) customer logs on to AOL and unread e-mail has been stored in the account. See America Onlinev. AT&T Corp., 243 F.3d 812, 815 (4th Cir. 2001) (“[I]n connection with its e-mail service, AOLadvises its subscribers that they have received e-mail by displaying the words ‘You Have Mail,’ byplaying a recording that announces, ‘You’ve got mail,’ and by depicting an icon of a traditional mailboxwith the red flag raised.”).

33. See Ex Parte Jackson, 96 U.S. 727, 732 (1877); 2 W. LAFAVE, SEARCH & SEIZURE § 3.7(c), at 362(3d ed. 1996).


lent of postal mail, the officer will conclude that the Fourth Amendmentrequires him to obtain a warrant before he can access the e-mail.34

Imagine that the second police office approaches the same problem from anexternal perspective. To him, the facts look quite different. Looking at how theInternet actually works, the second police officer sees that when A sent thee-mail to B, A was instructing his computer to send a message to his InternetService Provider (ISP) directing the ISP to forward a text message to B’s ISP.35

To simplify matters, let’s say that A’s ISP is EarthLink, and B’s ISP is AmericaOnline (AOL). EarthLink’s computers received A’s instructions, copied the textmessage, and then sent out another copy in the direction of the AOL server. Thate-mail crossed the Internet until it arrived at the AOL mail server, whichhappens to be located in Virginia.36 The next morning, when B sat at his deskand clicked on the icon to read the message from A, B was instructing hiscomputer to send a request to the AOL server to run off a copy of the messageand send it to him at his desk.

From the second officer’s external perspective, obtaining the e-mail seemsquite different from how it looked to the first officer. The second officer seesthat he can obtain a copy of the e-mail from any one of four sources: A, whosent the e-mail; the Earth Link server located in California, which kept a copybefore sending another copy to AOL; the AOL server in Virginia, whichretained a copy in B’s account; or B, who received a copy when he logged onand read the e-mail. To avoid tipping off A or B, the officer will probably wantto go to the system administrator at EarthLink or AOL to get a copy of themessage straight from their computers.

What process does the Fourth Amendment require? The second officer willreason that A sent a copy of the e-mail communication to a third party (theEarthLink computer), disclosing the communication to the third party andinstructing it to send the communication to yet another third party (AOL). Theofficer will ask, what process does the Fourth Amendment require to obtaininformation that has been disclosed to a third party and is in the third party’spossession? The officer will look in a Fourth Amendment treatise and locate tothe black letter rule that the Fourth Amendment permits the government to

34. Cf. United States v. Maxwell, 45 M.J. 406, 418 (C.A.A.F. 1996) (“In a sense, e-mail is like aletter. It is sent and lies sealed in the computer until the recipient opens his or her computer andretrieves the transmission. The sender enjoys a reasonable expectation that the initial transmission willnot be intercepted by the police.”).

35. This process works as follows:

When a user sends mail, she normally composes the message on her own computer, thensends it off to her ISP’s mail server. At this point her computer is finished with the job, but themail server still has to deliver the message. It does this by finding the recipient’s mail server,talking to that server and delivering the message. It then sits on that second mail server untilthe recipient comes along to read his mail, when he retrieves it onto his own computer.

Ken Lucke, Reading E-mail Headers, at http://www.stopspam.org/email/headers/headers.html (1997).36. See Maxwell, 45 M.J. at 416 (noting that AOL e-mails “are privately stored for retrieval on

AOL’s centralized and privately-owned computer bank located in Vienna, Virginia.”).


obtain information disclosed to a third party using a mere subpoena.37 Theofficer can simply subpoena the system administrator to compel him to producethe e-mails. No search warrant is required.

Who is right? The first officer or the second? The answer depends on whetheryou approach the Internet from an internal or external perspective. From aninternal perspective, the officers need a search warrant; from the externalperspective, they do not.38

2. Do Search Warrants Allow Remote Network Searches?

Let’s consider a second example, one that reverses the implications of theinternal and external approaches. Imagine that our two police officers give upon e-mail conspiracies and instead start investigating a local business that is afront for the mob in New York. The officers learn that the mob has stored a fullset of records of the mob’s illegal activities on the business’s computer network.The officers obtain a search warrant to search the New York office of thebusiness for the computer files. Importantly, the Fourth Amendment requires thewarrant to be fairly narrow; the warrant must specifically name the place thatwill be searched (“the business offices of the Mobfront Company, 123 PineStreet, Suite 200”) and name the evidence that will be seized (“computer filescontaining evidence of organized crime activity”).39 The warrant gives theofficers a limited grant of authority: It allows them to search the precise locationof the business for the precise evidence described, and no more.40

Imagine that when the officers execute the search, they find several computerterminals inside the business offices that are connected to the network, but theycannot find the central computer server that stores the network’s files. In fact,the network server is located hundreds or thousands of miles away, in anotherstate, or perhaps even another country. The officers will face a question: Does

37. See Donovan v. Lone Steer, Inc., 464 U.S. 408, 414–15 (1984); United States v. Dionisio, 410U.S. 1, 8 (1973); 2 WAYNE R. LAFAVE, SEARCH AND SEIZURE § 4.13(a), at 726 (3d ed. 1996).

38. It is possible to reconcile these two approaches, at least in part. Although the case law on thisissue is sparse, the Fourth Amendment rules governing subpoenas apply even when information isprotected by a “reasonable expectation of privacy.” See In re Subpoena Duces Tecum, 228 F.3d 341,348–49 (4th Cir. 2000). As a result, even though a search warrant would be required to access certaindocuments if the government conducted the search directly, the government can alternatively use asubpoena to obtain the materials without a warrant. See id. This rule applies to postal mail, as well. SeeUnited States v. Barr, 605 F. Supp. 114, 119 (S.D.N.Y. 1985). However, this does not mean that theinternal and external approaches yield the same result. If the sending of an e-mail to an ISP is adisclosure that eliminates a reasonable expectation of privacy (the external result), then the governmentcould access the e-mail themselves or use a defective subpoena without violating the Fourth Amend-ment.

39. See U.S. CONST. amend. IV. (“[N]o Warrants shall issue, but upon probable cause, supported byOath or affirmation, and particularly describing the place to be searched, and the persons or things to beseized.”).

40. See Maryland v. Garrison, 480 U.S. 79, 84 (1987) (“By limiting the authorization to search to thespecific areas and things for which there is probable cause to search, the requirement ensures that thesearch will be carefully tailored to its justifications, and will not take on the character of thewide-ranging exploratory searches the Framers intended to prohibit.”).


their search warrant allow them to search the terminals inside the business andretrieve the information stored remotely on the network?

The first police officer, who prefers an internal perspective, will say “yes.”This officer will approach the terminal and see various icons indicating thepresence of the network’s files. To him, the files listed are virtually presentinside the terminal; he can access them from the network exactly as he would beable to access files stored on a local hard drive or floppy diskette. The firstofficer will look at the warrant, see that it authorizes him to search “123 PineStreet, Suite 200” for “computer files containing evidence of organized crimeactivity,” and will conclude that the warrant authorizes him to search theterminal for the evidence.

The second police officer, who approaches the same problem from an exter-nal perspective, will disagree. He will reason that if he sits down at the terminaland starts looking through the files on the network, he will actually be instruct-ing the terminal to send commands to the remote central server to run off copiesof the files and send them back to him. Retrieving the files will not search “123Pine Street, Suite 200,” as his warrant allows, but rather will direct a search ofthe physical location of the server located hundreds or even thousands of milesaway. If the server is located in California, he will be searching a place inCalifornia; if the server is in Canada, he will be searching a place in Canada.41

The search warrant does not allow the officer to execute a search outside of“123 Pine Street, Suite 200,” however, much less in another state or a foreigncountry. Accordingly, the second officer will conclude that the search warrantdoes not allow him to search the terminal for the network’s remotely storedfiles. From an internal perspective, the search warrant authorizes the officers tosearch the terminal; from an external perspective, it does not.

Notably, the privacy implications of the two perspectives in this secondhypothetical reverse the implications of the first one. In the e-mail hypothetical,the internal perspective offers more privacy protection, whereas the opposite istrue in the case of the remote network search. This should not surprise us. Theinternal and external perspectives offer two distinct and unrelated set of theInternet’s facts, and the legal implications of the two perspectives should varydepending on the issue. In some cases the internal perspective will offer factsthat lead to greater privacy protection, but in other cases the external perspec-tive will be more protective.

B. INTERNET GOVERNANCE

Internet governance has received a great deal of attention among scholarlycommentators.42 We can define Internet governance as the study of how law,

41. See Patricia L. Bellia, Chasing Bits Across Borders, 2001 U. CHI. LEGAL F. 35, 42 (consideringthe Fourth Amendment implications of cross-border searches).

42. See, e.g., Henry H. Perritt, Jr., Towards a Hybrid Regulatory Scheme for the Internet, 2001 U.CHI. LEGAL F. 215; Lawrence Lessig, Open Code and Open Societies: Values of Internet Governance,


legal institutions, and computer code collectively regulate and define the virtualworld of cyberspace.43 Here, as with the Fourth Amendment, approaches toInternet governance are influenced heavily by the competing views of theinternal and external approaches. This section explains how the internal-external critique explains competing views of Internet governance. It beginswith Professor Lawrence Lessig’s insight that “code is law”; turns next to a1996 federal court decision, Cyber Promotions v. America Online44; and con-cludes with a broader discussion of how the internal and external perspectivesshed light on theories of Internet governance.

1. Code is Law—Or Is It?

Professor Lawrence Lessig is today’s most influential cyberlaw theorist.Lessig has popularized the idea that “code is law.”45 According to this idea, thehardware and software code that make up the Internet regulates the virtualworld of cyberspace much like law regulates the physical world.46 The natureand values of cyberspace hinge on the coding decisions of programmers, whichmeans that the decisions made by big companies like America Online andMicrosoft have the force of law in cyberspace.

Lessig argues that because code is law, we should apply the legal constructsto code in cyberspace that we normally apply to law in the physical world. Forlawyers, this approach is striking primarily because it assigns the U.S. Constitu-tion an essential role in regulating the Internet’s computer hardware and soft-ware.47 If companies like AOL and Microsoft construct “the functional equivalentof law”48 in cyberspace, Lessig argues, their decisions should be just as subjectto constitutional scrutiny as would law in the physical world. To do otherwisewould make cyberspace a remarkable anomaly, a space totally unregulated bythe Constitution.49 Lessig urges us to rethink the state action doctrine to

74 CHI.-KENT. L. REV. 1405 (1999). The notion of “Internet governance” as a distinct field withinInternet law was recently acknowledged by the Duke Law Journal, which held its thirtieth annualadministrative law issue on “Governance of the Internet.” See James Boyle, A Nondelegation Doctrinefor the Digital Age?, 50 DUKE L.J. 5 (2000) (introducing the issue).

43. See generally Boyle, supra note 42.44. 948 F. Supp. 436 (E.D. Pa. 1996).45. See LESSIG, supra note 20, at 6. Lessig credits this idea largely to William Mitchell as well as to

others. See id. at 6 n.7 (citing WILLIAM J. MITCHELL, CITY OF BITS: SPACE, PLACE, AND THE INFOBAHN 111(1995)).

46. See id. at 6.47. See id. (“In speaking of a constitution in cyberspace, we are simply asking: What values are

protected there? What values will we build [by code] into the space to encourage certain forms oflife?”).

48. Id. at 217.49. See id. As Lessig puts it: “If code functions as law, then we are creating the most significant new

jurisdiction since the Louisiana Purchase, yet were are doing it just outside the Constitution’s review.”See id.


accomplish these goals.50 According to Lessig, we should recognize that thecodemakers such as Microsoft and AOL act as virtual governments that exercisereal control over the virtual world of cyberspace and consider subjecting theirdecisions to constitutional scrutiny.51

Viewing Lessig’s theories through the lens of the internal-external critiqueproves quite revealing, I think. The phrase “code is law” expresses a relation-ship between the internal and external perspectives. “Code is law” means thatcode from the external perspective has the force of law from the internalperspective. The programmer’s code defines the architecture of the virtual worldthat a user encounters. Because external code is internal law, we should regulateexternal code from an internal perspective.

This internal focus also seems to provide the key to understanding Lessig’stheory of applying the Constitution in cyberspace. Lessig is more eager thanmost to apply the Constitution to the Internet from an internal perspective.52 Heurges us to apply the Constitution to the world the Internet user perceives, ratherthan to the bricks-and-mortar world that exists off-line. State action offers aclear example. To answer who is a state actor, Lessig suggests, we should lookat the world from an Internet user’s perspective and determine who has powersthat resemble those of the government.53 We can translate the values of theConstitution into cyberspace only by recognizing the Internet user’s world asthe functional equivalent of the physical world and trying to match the rules ofrealspace to the rules of cyberspace.

Lessig has developed a complex explanation for why we should endorse thisinternal perspective. According to Lessig, fidelity to the Constitution requires usto “translate”54 our “constitutional values”55 from real space to cyberspace toaddress “latent ambiguities”56 in our constitutional structure. I am not a constitu-tional theorist, and I will leave it to others to critique this approach.57 However,

50. Lessig suggests the need for “a revolution in American constitutional law for the Court . . . tomove beyond the limits of state action” to be “more faithful” to our constitutional traditions. Id. at217–18.

51. See id. at 217.52. Given the richness and complexity of Professor Lessig’s works, viewing his scholarship through

this single lens necessarily ignores a great deal of subtlety. As a first-order approximation, however, Ithink it captures some of the unique flavor of Professor Lessig’s ideas.

53. LESSIG, supra note 20, at 217–18. Thus Professor Lessig refers frequently to the importance of“constitutional values.” See, e.g., Lawrence Lessig, What Things Regulate Speech: CDA 2.0 v.Filtering, 38 JURIMETRICS J. 629, 664 (1998). Lessig does not define what “constitutional values” are,but in his work they appear to be values that from an internal perspective would be something that theConstitution may address if they occurred in the physical world.

54. LESSIG, supra note 20, at 114 (referring to the strategy of “finding a current reading of theoriginal Constitution that preserves its original meaning in the present context.”).

55. Lessig, supra note 53, at 664.56. LESSIG, supra note 20, at 217 (defining “latent ambiguity” as a question that the Framers did not

answer because it was not raised in their day).57. Admittedly, I have expressed some skepticism of it before. See Orin S. Kerr, The Fourth

Amendment in Cyberspace: Can Encryption Create a “Reasonable Expectation of Privacy”?, 33 CONN.L. REV 503, 525 n.108 (2001).


for our purposes the key is to note that Lessig’s theory compares to moretraditional approaches primarily in its willingness to adopt a more internalapproach to constitutional law, in which the user’s perspective matters.58

2. Cyber Promotions v. America Online

The importance of internal and external approaches to the area of Internetgovernance is not merely a question of theory. The internal and externalperspectives have clashed in one case that squarely addressed the application ofthe state action doctrine to the Internet: Cyber Promotions v. America Online.59

The Cyber Promotions case arose from a dispute over “spam,” unsolicitedcommercial e-mail. Cyber Promotions was a spamming company that violatedAOL’s antispam policy; in response, AOL banned Cyber Promotions from itsnetwork. Cyber Promotions filed suit in federal court, alleging that AOL’s banviolated Cyber Promotions’s First Amendment rights to speak to AOL’s subscrib-ers. To succeed on this claim, Cyber Promotions first had to establish that AOLwas a state actor governed by the First Amendment, rather than merely a privateentity.

Cyber Promotions tried to do this by offering an internal perspective. To anInternet user, its lawyers argued, decisions by America Online “ha[ve] thecharacter of state action.”60 To users, AOL acts like the government. It providesa virtual town square where “public discourse, conversations and commercialtransactions can and do take place,”61 just like in a real town square. BecauseAOL took on this governmental role from the perspective of a user, AOL shouldbe considered a state actor.62

The district court rejected this argument by applying an external perspective.First, the court framed its analysis by describing the Internet as “a global Webof linked networks and computers . . . owned and managed by private entitiesand persons.”63 Note the external description, focusing on the physical hard-ware of the computers attached to the Internet rather than the virtual world itcan help create. Within this external construct, AOL was simply a Virginiacompany that provided Internet access to its customers and was not owned in

58. Professor Tribe has suggested an internal approach to the Constitution (and in particular theFourth Amendment) in his essay on the Constitution in Cyberspace. See LAWRENCE H. TRIBE, THE

CONSTITUTION IN CYBERSPACE, at http://www.eff.org/Legal/cyber_constitution.paper (1991). “Axiom 5”of Tribe’s essay urges the courts to approach the Constitution in cyberspace in a technology-neutralway. As he puts it, “Constitutional Principles Should Not *Vary With Accidents of Technology*.” Id.According to Tribe, courts should treat the Internet’s new means of communicating just like theirphysical equivalents. This appears to reflect an internal approach, or at least a Constitution that followsan internal standard: Tribe suggests that courts focus on the role that new technologies play in theexperience of their users, whereas physical reality should be seen as mere “accidents of technology.” Id.

59. 948 F. Supp. 436 (E.D. Pa. 1996).60. Id. at 441.61. Id. at 442.62. Id.63. Id. at 441.


whole or in part by the government.64 The court instructed that “AOL is merelyone of many private online companies which allow its members access to theInternet through its e-mail system where they can exchange information withthe general public. The State has absolutely no interest in, and does not regulate,this exchange of information . . . around the world.” 65

Accordingly, AOL was not a state actor, and its decision to ban a spammercould not violate the First Amendment. That an Internet user may think of AOLas the functional equivalent of government in cyberspace was irrelevant becausethe company itself was plainly a private entity and not a state actor. Whatmattered was the external question of corporate identity, not the internal ques-tion of users’ perceptions.

3. Internet Governance from an External Perspective

That questions of Internet governance divide sharply along internal andexternal lines should not surprise us. The very concept of “Internet governance”is an internal construct. It seeks to analogize the creation of rules in the physicalworld (traditional questions of governance) to the creation of rules in cyber-space (Internet governance). Whether this enterprise makes any sense dependsupon whether one approaches the Internet internally or externally.

From an external perspective, the analogy appears nonsensical. AOL governscyberspace from the perspective of its subscribers only in the way that thereferee of a high school football game governs the game from the perspective ofits players, or Steven Spielberg governs his movies from the perspective of theirviewers. AOL calls the shots, but that does not make it the government. Thepower to control a user’s experience is an important power, but its importancedoes not transform it into a governmental power. To an external viewer, viewingAOL as a state actor in cyberspace makes as much sense as calling the referee astate actor in “footballspace,” or Steven Spielberg a state actor in “moviespace.”The problem derives from the maxim “code is law.” Saying that the power ofcode is akin to the power of law is simply too loose a use of the word “law” tobe helpful. If code is law to an Internet user, then a sports referee’s calls are lawto an athlete, and Steven Spielberg’s decisions about how to shoot a movie arelaw to a movie viewer. The internal approach simply misses the critical differ-ence between reality and virtual reality.66

Conversely, from an internal perspective the external approach fails to trans-

64. See id.65. Id. at 442.66. Perhaps the most remarkable example of this within Lessig’s work is his extensive discussion of

a “rape in cyberspace.” See LESSIG, supra note 20, at 74–78. Lessig recounts how a person whocontrolled a character named “Bungle” in an online game called “LamdaMOO” was able to manipulatethe game such that the character “raped” several other characters in the game. See id. at 74–75.Although Lessig recognizes that no real woman was raped, he insists that “all will see a link betweenrape and what happened to these women here. Bungle used his power over these women for his own(and against their) sexual desire; he sexualized his violence and denied them even the dignity ofregistering their protest.” Id. at 75. While it is possible to envision a “link,” the link must be highly


late the values of the real world into cyberspace.67 The internal viewer will tendto see the external observer as rigid, myopic, and lacking in imaginationbecause the external perspective does not recognize the importance of theinternal world.68

C. COMPUTER CRIME

Internal and external approaches have also influenced the law of computercrime.69 Although some computer crimes present new forms of criminality,many are simply traditional, preexisting crimes committed using the Internet. Touse the cliche, they are “old wine in new bottles.”70 Examples include onlinefrauds,71 cyberstalking,72 Internet gambling,73 and offenses involving the onlinedistribution of child pornography.74

Prosecutions of these crimes require the courts to apply preexisting criminallaws to the Internet, resulting in a series of clashes between the internal andexternal perspectives. This section contrasts two recent federal court of appealsdecisions in which defendants tried and failed to overturn their convictions byinvoking differing perspectives. In the first case, the defendant offered aninternal view, but the court adopted an external view; in the second, thedefendant offered an external view, but the court endorsed an internal view. Thesection concludes by analyzing a rare case in which the internal and externalperspectives produced similar (although not identical) results: the SupremeCourt’s First Amendment decision striking down the Communications DecencyAct in Reno v. American Civil Liberties Union.75

1. United States v. Kammersell

The Tenth Circuit’s decision in United States v. Kammersell76 provides adramatic example of the importance of choosing between an internal andexternal perspective of the Internet. In this case, nineteen-year-old Matthew

tenuous at best: It is the link between a brutal rape and a fictional story of a brutal rape. Surely thedifference is more striking than any similarity.

67. See id. at 217–18. See also Paul Schiff Berman, Cyberspace and the State Action Debate: TheCultural Value of Applying Constitutional Norms to “Private” Regulation, 71 U. COLO. L. REV. 1263,1270 (2000) (“[W]hether America Online is public or private, there are certain values that we hold as acommunity, values that America Online may be threatening.”).

68. See LESSIG, supra note 20, at 217–18.69. See generally Scott Charney & Kent Alexander, Computer Crime, 45 EMORY L. J. 931 (1996).70. See United States v. LaMacchia, 871 F. Supp. 535, 536 (D. Mass. 1994); see Michael Edmund

O’Neill, Old Crimes in New Bottles: Sanctioning Cybercrime, 9 GEO. MASON L. REV. 237, 237 (2001).71. See, e.g., Gretchen Morgenson, S.E.C. Says Teenager Had After-School Hobby: Online Stock

Fraud, N.Y. TIMES, Sept. 21, 2000, at A1.72. See U.S. DEP’T OF JUSTICE, CYBERSTALKING: A NEW CHALLENGE FOR LAW ENFORCEMENT AND

INDUSTRY (1999), http://www.usdoj.gov/criminal/cybercrime/cyberstalking.htm.73. See, e.g., United States v. Cohen, 260 F.3d 68, 70–71 (2d Cir. 2001).74. See generally PHILIP JENKINS, BEYOND TOLERANCE: CHILD PORNOGRAPHY ON THE INTERNET (2001).75. 521 U.S. 844 (1997).76. 196 F.3d 1137 (10th Cir. 1999).


Kammersell used America Online’s “instant message” service to send a bombthreat over the Internet from Riverdale, Utah (a suburb of Ogden, Utah) to hisgirlfriend’s computer at work in downtown Ogden, a few miles away.77 Thegovernment prosecuted Kammersell under 18 U.S.C. § 875(c), which makes it afederal felony to send an interstate communication “containing . . . any threat toinjure the person of another.”78

The question facing the Court was whether Kammersell’s instant messageconstituted an “interstate” threat. Kammersell argued that it did not. He offeredan internal account of his conduct. From his perspective, he was located in Utahand had sent the threat to his girlfriend in Utah. His girlfriend was just a fewmiles away, in the same state. From Kammersell’s perspective, there wasnothing interstate about his threat, and the absence of a federal interstate nexusrequired the court to vacate his conviction.79

The government countered by approaching Kammersell’s threat from anexternal perspective. Because America Online’s servers are located in Virginia,the government noted, every AOL instant message must be routed from its pointof origin to AOL’s servers in Virginia, and then on to its destination. Unbe-knownst to Kammersell, his instant message had traveled from Utah to Virginia,and then back to Utah. Kammersell’s threat was in fact an “interstate” threat, asit had traveled most of the way across the country twice in the course of beingdelivered.80

Did Matthew Kammersell send an interstate threat? From an internal perspec-tive, no; from an external perspective, yes. The Tenth Circuit adopted thegovernment’s external perspective and affirmed the conviction.81

2. United States v. Thomas

A 1996 obscenity case decided by the Sixth Circuit, United States v. Thom-as,82 provides a mirror image of Kammersell. As in Kammersell, the defendant’s

77. See id. at 1138. Instant messages are private communications between two Internet accounts inuse that are delivered immediately to the other. See America Online v. AT&T Corp., 64 F. Supp. 2d 549,566 (E.D. Va. 1999), aff’d in part, vacated in part, 243 F.3d 812 (4th Cir. 2001).

78. 18 U.S.C. § 875(c) (2000) states in full: “Whoever transmits in interstate or foreign commerceany communication containing any threat to kidnap any person or any threat to injure the person ofanother, shall be fined under this title or imprisoned not more than five years, or both.” The statute hasbeen applied fairly often to cases involving Internet threats. See, e.g., United States v. Morales, 272F.3d 284, 285 (5th Cir. 2001); United States v. Alkhabaz, 104 F.3d 1492, 1493 (6th Cir. 1997).

79. See Kammersell, 196 F.3d at 1139. Notably, this is not an entirely internal account; it considersthe location of Kammersell and his girlfriend in the physical world. However, the account is internalfrom the standpoint of whether his communication traveled across state lines.

80. See id.81. See id. at 1140.82. 74 F.3d 701 (6th Cir. 1996). The Thomas case received significant attention at the time primarily

for its interesting First Amendment implications. See generally Jennifer K. Michael, Note, Where’s“The Nastiest Place on Earth?” From Roth to Cyberspace, or, Whose Community Is It, Anyway? TheUnited States Court of Appeals for the Sixth Circuit Addresses Local Community Standards in UnitedStates v. Thomas, 30 CREIGHTON L. REV. 1405 (1997); Stephen G. Kabalka, Case Note, Application of


liability in Thomas hinged on whether the facts were approached internally orexternally. Unlike Kammersell, however, the defendant in Thomas asked thecourt to apply an external perspective, and the government advocated aninternal perspective.

Robert and Carleen Thomas operated a computer bulletin board service fromtheir home in California starting in 1991.83 The bulletin board billed itself as“The Nastiest Place on Earth” and provided its paying customers with hard-corepornography.84 Officials in Memphis, Tennessee began an investigation intowhether the Thomases had violated federal obscenity laws. Eventually they setup an operation in which an undercover postal inspector in Memphis opened anaccount and used a computer to receive digital photographs formatted as GIFfiles from the Thomases’ server.85 The U.S. Attorney in Memphis charged theThomases with almost a dozen violations of federal law, among them violationsof 18 U.S.C. § 1465.86 This statute makes it a federal crime to use a means ofinterstate commerce to transport an obscene “book, pamphlet, picture, film . . .[or] image” in interstate commerce.87 The government’s theory was that theThomases had used the Internet (a means of interstate commerce) to transportthe GIF files (the images) in interstate commerce (from the server in Californiato the postal inspector in Tennessee). The jury convicted the Thomases on mostof the counts, including six counts of violating 18 U.S.C. § 1465.88

On appeal, the Thomases argued that they had not violated the statute. Theyoffered an external account of their conduct. They had not sent “images” to theundercover officer in Memphis, they reasoned, but merely had sent a “string of0’s and 1’s”89 from one computer to another. The fact that the “string of 0’s and1’s” was a GIF file that could be translated by the postal inspector’s computerinto a visual image did not mean that the 0s and 1s were an “image” covered by

Existing Obscenity Laws to Computer Transmissions United States v. Thomas, 74 F.3d 701 (6th Cir.1996), 64 TENN. L. REV. 215 (1996).

83. Thomas, 74 F.3d at 704.84. Id. at 705.85. Id. GIF stands for “Graphic Interchange Format,” and is a popular file format for electronic files

that can be read by a computer as a picture. See United States v. Lacy, 119 F.3d 742, 745 n.2 (9th Cir.1997).

86. 18 U.S.C. § 1465 (2000).87. The statute provides:

Whoever knowingly transports or travels in, or uses a facility or means of, interstate or foreigncommerce or an interactive computer service . . . in or affecting such commerce for thepurpose of sale or distribution of any obscene, lewd, lascivious, or filthy book, pamphlet,picture, film, paper, letter, writing, print, silhouette, drawing, figure, image, cast, phonographrecording, electrical transcription or other article capable of producing sound or any othermatter of indecent or immoral character, shall be fined under this title or imprisoned not morethan five years, or both.

Id.88. Thomas, 74 F.3d at 706.89. Id. at 706–07.


the federal obscenity law.90

The Sixth Circuit rejected this external argument in favor of an internaldescription of the Thomases’ conduct. According to the court, “the means bywhich the GIF files were transferred” did not matter so long as “the transmis-sions began with computer-generated images in California and ended with thesame computer-generated images in Tennessee.”91 Because the data sent by theThomases appeared as an image when “viewed on a computer screen”92 by acomputer user, it was an “image” according to federal obscenity law. From thecourt’s internal perspective, technical details such as “[t]he manner in which theimages moved”93 seemed irrelevant. The court affirmed the conviction.94

3. Internal and External Aligned: Reno v. American Civil Liberties Union

In both Kammersell and Thomas, the courts’ outcomes depended on theirperspectives. The internal and external approaches do not always produceopposing results, however. An interesting case in which the two perspectives ledto similar outcomes was the Supreme Court’s 1997 decision in Reno v. Ameri-can Civil Liberties Union.95

Reno raised a First Amendment challenge to the 1996 CommunicationsDecency Act (CDA), a law that Congress enacted to protect minors fromharmful material on the Internet.96 The challenge focused on two provisions:one provision that prohibited the knowing transmission “by means of a telecom-munications device”97 of obscene or indecent messages to any recipient who isa minor98 and another that prohibited using an “interactive computer service”99

to knowingly send or display patently offensive messages in a manner that isavailable to a minor.100

90. Id. at 707.91. Id.92. Id.93. Id.94. Id. at 706.95. 521 U.S. 844 (1997).96. 47 U.S.C. § 223 (2000).97. 47 U.S.C. § 223(a).98. Id. The relevant text of this provision stated:

Whoever . . . by means of a telecommunications device knowingly . . . makes, creates, orsolicits, and initiates the transmission of, any comment, request, suggestion, proposal, image,or other communication which is obscene or indecent, knowing that the recipient of thecommunication is under 18 years of age, regardless of whether the maker of such communica-tion placed the call or initiated the communication . . . shall be fined under title 18, UnitedStates Code, or imprisoned not more than two years, or both.

Id.99. 47 U.S.C. § 223(d).100. Id. The relevant text of this provision stated:

Whoever . . . knowingly uses an interactive computer service to send to a specific person orpersons under 18 years of age, or uses any interactive computer service to display in a manneravailable to a person under 18 years of age, any comment, request, suggestion, proposal,


The majority opinion by Justice Stevens held that both provisions violatedthe First Amendment.101 The Court applied an external perspective. To themajority, the Internet was a worldwide network of computers that allowed itsusers to access resources via host computers.102 Although the Court recognizedthat an internal perspective exists—at one point the opinion notes that “from thereaders’ viewpoint,” the World Wide Web resembles “a vast library . . . and asprawling mall”103—the majority viewed the CDA as a speech restriction on theuse of a specific network technology akin to a restriction on television orradio.104 From this external perspective, the CDA could not survive constitu-tional scrutiny because less restrictive technological means existed by whichCongress could have achieved the CDA’s legitimate goals without burdeningspeech.105

Justice O’Connor, joined by Chief Justice Rehnquist, authored a partialconcurrence and partial dissent.106 Justice O’Connor approached the CDA froman internal perspective. According to Justice O’Connor, the CDA was “littlemore than an attempt by Congress to create ‘adult zones’ on the Internet.”107

Citing Professor Lessig, Justice O’Connor argued that the CDA was an attemptto regulate cyberspace much like a zoning law attempts to regulate the physicalworld.108 Its constitutionality depended on whether the CDA’s attempt to zonechat rooms and others areas of cyberspace satisfied the First Amendmentstandard that the Court had developed to test efforts to zone the physicalworld.109 Applying this standard to an internal picture of the Internet, JusticeO’Connor agreed with the external majority view that parts of the CDA didviolate the First Amendment, but would have held that other parts survived FirstAmendment scrutiny.110 Because the CDA raised similar problems from an

image, or other communication that, in context, depicts or describes, in terms patentlyoffensive as measured by contemporary community standards, sexual or excretory activities ororgans, regardless of whether the user of such service placed the call or initiated thecommunication; or knowingly permits any telecommunications facility under such person’scontrol to be used for an activity prohibited by paragraph with the intent that it be used forsuch activity, shall be fined under title 18, United States Code, or imprisoned not more thantwo years, or both.

Id.101. See Reno v. Am. Civil Liberties Union, 521 U.S. 844, 864–85 (1997).102. See id. at 849–51.103. Id. at 853.104. See id. at 868–70.105. See id. at 879.106. See id. at 886 (O’Connor, J., concurring in the judgment in part and dissenting in part).107. Id. (O’Connor, J., concurring in the judgment in part and dissenting in part).108. See id. at 889 (O’Connor, J., concurring in the judgment in part and dissenting in part) (citing

Lawrence Lessig, Reading the Constitution in Cyberspace, 45 EMORY L.J. 869, 886 (1996)).109. See id. at 886 (O’Connor, J., concurring in the judgment in part and dissenting in part)

(contending that “portions of the CDA are unconstitutional because they stray from the blueprint ourprior cases have developed for constructing a ‘zoning law’ that passes constitutional muster.”).

110. See id. at 897 (O’Connor, J., concurring in the judgment in part and dissenting in part). JusticeO’Connor concluded that


internal perspective that it raised from the majority’s external perspective,Justice O’Connor reached a similar (although not identical) result as the exter-nal majority.

D. COPYRIGHT LAW AND THE INTERNET

Copyright law provides a final example of how internal and external perspec-tives shape Internet law. Copyright law grants authors rights to control thereproduction of their original works that are fixed in a tangible medium ofexpression.111 How copyright laws should apply to the Internet has provenhighly controversial.112 Here, too, competing internal and external viewpointsinfluence the shape of the law.

We can see how perspectives affect copyright law on the Internet by studyingthe recent lawsuit against the Internet service MP3.com. In January 2000,MP3.com began offering Internet users a service that allowed them to accesstheir compact disc (CD) collections from anywhere in the world via theInternet.113 To enjoy the service, users needed to register and establish that theyowned a particular collection of CDs. A user could do this in two ways: first “byinserting his copy of the commercial CD into his computer CD-Rom drive for afew seconds (the ‘Beam-it Service’),”114 or second by “purchas[ing] the CDfrom one of defendant’s cooperating online retailers (the ‘Instant ListeningService’).”115 Registered users could then log on to mymp3.com and requestspecific songs, which MP3.com would provide in the form of compressed

the constitutionality of the CDA as a zoning law hinges on the extent to which it substantiallyinterferes with the First Amendment rights of adults. Because the rights of adults are infringedonly by the ‘display’ provision and by the ‘indecency transmission’ and ‘specific person’provisions as applied to communications involving more than one adult, I would invalidate theCDA only to that extent. Insofar as the ‘indecency transmission’ and ‘specific person’provisions prohibit the use of indecent speech in communications between an adult and one ormore minors, however, they can and should be sustained.

Id. at 896 (O’Connor, J., concurring in the judgment in part and dissenting in part).111. See 17 U.S.C. § 102(a) (2000) (“Copyright protection subsists, in accordance with this title, in

original works of authorship fixed in any tangible medium of expression, now known or laterdeveloped, from which they can be perceived, reproduced, or otherwise communicated, either directlyor with the aid of a machine or device.”). For a comprehensive treatment, see MELVILLE B. NIMMER AND

DAVID NIMMER, NIMMER ON COPYRIGHT (1999). For an engaging treatment of the history of copyrightlaw, see BENJAMIN KAPLAN, AN UNHURRIED VIEW OF COPYRIGHT (1967).

112. See generally LAWRENCE LESSIG, THE FUTURE OF IDEAS (2001); Jessica Litman, RevisingCopyright Law for the Information Age, 75 OR. L. REV. 19 (1996); Chip Patterson, Note, CopyrightMisuse and Modified Copyleft: New Solutions to the Challenges of Internet Standardization, 98 MICH.L. REV. 1351 (2000); Alfred C. Yen, A Personal Injury Law Perspective on Copyright in an InternetAge, 52 HASTINGS L.J. 929 (2001).

113. See UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349, 350 (S.D.N.Y. 2000) (Rakoff,J.).

114. Id.115. Id.


mp3-formatted files.116 The owners of MP3.com provided this service bybuying tens of thousands of copyrighted compact discs, and then copying themonto MP3.com’s servers. Whenever a registered user would request a particularsong, MP3.com’s servers would run off a copy of the song from their serversand distribute it directly to the user over the Internet.117

The five major record companies did not appreciate this arrangement andsued MP3.com for copyright infringement.118 Did the service violate copyrightlaw? The record companies relied on an external perspective to argue that theanswer was “yes.” To the record companies, MP3.com first made unauthorizedcopies by copying the record companies’ copyrighted discs on to MP3.com’sservers and then made and distributed additional unauthorized copies whenevera user requested a track or entire disc. From this external perspective, MP3.comhad twice copied materials without authorization from the copyright holders.Accordingly, MP3.com plainly infringed upon the record companies’ copy-rights.

MP3.com responded to this charge with an internal perspective of the sametransaction. From the perspective of an Internet user, its lawyers argued, MP3.comsimply allowed legitimate owners of music to access their CD collectionwithout physically having to carry their CDs.119 MP3.com provided a virtualstorage locker that permitted users to do in cyberspace what the physical CDspermitted them to do in realspace: listen to music they had legitimately pur-chased. Accordingly, MP3.com argued that there was no actual infringement ofthe record companies’ copyrights, and that if there was infringement, it wasmerely a de minimis “space shift” from realspace to cyberspace permitted underthe fair use doctrine.120

Did MP3.com’s service violate the copyright laws? Once again, the choice ofperspective helps determine the answer. From an external perspective, probablyyes; from an internal perspective, probably no. The district court adopted theexternal perspective and ruled in favor the record companies.121

116.

MPEG-1 Audio Layer 3 (commonly known as ‘MP3’) is the most popular digital audiocompression algorithm in use on the Internet, and the compression it provides makes an audiofile ‘smaller’ by a factor of twelve to one without significantly reducing sound quality. MP3’spopularity is due in large part to the fact that it is a standard, non-proprietary compressionalgorithm freely available for use by anyone, unlike various proprietary (and copyright-secure) competitor algorithms.

Recording Indus. Ass’n of America v. Diamond Multimedia Sys. Inc., 180 F.3d 1072, 1074 (9th Cir.1999).

117. See UMG Recordings, 92 F. Supp. 2d. at 350.118. See id.; David Akin, Legions of Record Company Lawyers Gird for Battle with the Internet

Lawyers, NAT’L POST, Jan. 26, 2000, at C9.119. See UMG Recordings, 92 F. Supp. 2d. at 350 (“[The] defendant seeks to portray its service as

the ‘functional equivalent’ of storing its subscribers’ CDs”).120. See id. at 351; 17 U.S.C. § 117 (2000).121. See UMG Recordings, 92 F. Supp. 2d at 350 (“[A]lthough defendant seeks to portray its service

as the ‘functional equivalent’ of storing its subscribers’ CDs, in actuality defendant is re-playing for the


III. THE SIGNIFICANCE OF THE PROBLEM OF PERSPECTIVE

Many scholars believe that the field of Internet law offers nothing new.122

They reason that applying law to the Internet is like applying law to any new setof facts: draw analogies and then apply existing law.123 To these skeptics, thelaw of the Internet is merely “old wine in new bottles.”124 Although it has afancy New Economy name—“cyberlaw”125—there is nothing really new aboutthe law of the Internet.126

Cyberlaw’s proponents have offered only a partial response to these claims.Most prominently, Professor Lessig has defended the study of cyberlaw on theground that it can shed light on law beyond cyberspace.127 Studying cyberlawreveals how law interacts with other regulators of human conduct, such asarchitecture and social norms.128 This may be true, but it largely fails to addressthe skeptics’ criticism. If anything, Professor Lessig’s defense may fuel thecritics’ concerns: By defending cyberlaw based on its similarities to the rest oflaw, rather than its differences, the argument implicitly concedes that nothingnew happens when we apply law to the Internet.129

subscribers converted versions of the recordings it copied, without authorization, from plaintiffs’copyrighted CDs.”) From the court’s external perspective, MP3.com’s internal approach seemed “littlemore than a sham.” UMG Recordings, Inc., v. MP3.com, 56 U.S.P.Q.2d 1376, 1377 (S.D.N.Y. 2000).The court added:

Some of the evidence in this case strongly suggests that some companies operating in the areaof the Internet may have a misconception that, because their technology is somewhat novel,they are somehow immune from the ordinary applications of laws of the United States,including copyright law. They need to understand that the law’s domain knows no such limits.

Id. at 1379.122. See, e.g., Frank H. Easterbrook, Cyberspace and the Law of the Horse, 1996 U. CHI. LEGAL F.

207, 208 (criticizing the field of Internet law as merely being the “law of the horse,” a collection ofunrelated legal issues that happen to involve a technology); Christopher M. Kelly, The CyberspaceSeparatism Fallacy, 34 TEX. INT’L L. J. 413 passim (1999) (same); Joseph H. Sommer, AgainstCyberlaw, 15 BERKELEY TECH. L. J. 1145, 1147 (2000) (arguing that “‘cyberlaw’ and ‘the law of theInternet’ are not useful concepts,” and that cyberlaw is “nonexistent.”).

123. See Sommer, supra note 122, at 1148 (arguing that applying law to the Internet raises the sameissues as applying law to other sets of facts).

124. See United States v. LaMacchia, 871 F. Supp. 535, 536 (D. Mass. 1994).125. Lessig, supra note 4, at 501.126. See id. (reviewing the debate over intellectual property law in cyberspace and concluding that

“there is nothing new here. Cyberspace is another battleground for an old war.”).127. See Lawrence Lessig, supra note 4, at 502 (arguing that cyberlaw can teach “the limits on law

as a regulator and about the techniques for escaping those limits”). Other cyberlaw defenders havesuggested that cyberlaw’s identity derives from the way that its new facts may require amendingpreexisting doctrines. See, e.g., I. Trotter Hardy, The Proper Legal Regime for “Cyberspace”, 55 U.PITT. L. REV. 993, 1053–54 (1994) (contending that the Internet raises several new questions for thelegal system); Paul Schwartz, Privacy and Democracy in Cyberspace, 52 VAND. L. REV. 1609, 1612(1999) (arguing that the Internet’s facilitation of widespread information-sharing should prompt afederal law on fair information practices).

128. See Lessig, supra note 4, at 503–10.129. As cyberlaw critic Joseph Sommer has recognized:

Much of the best work [on cyberlaw] is not really ‘about’ cyberlaw at all. Instead, it appliesexisting legal doctrine or political theory to a new arena . . . . In applying our old law to


The problem of perspective suggests that the defenders of cyberlaw are right,although perhaps not for the reasons they have offered. Internet law does offersomething new—not so much in how we approach the law, but rather in the waythat we approach the facts. Modeling the Internet’s facts turns out to be lesssimple than the skeptics suggest, as it requires a choice between external andinternal constructions of those facts, between physical reality and virtual reality.Even accepting the skeptic’s view that Internet law is simply a matter ofapplying law to the facts, the Internet’s facts hinge on whether we approach theInternet internally or externally.

This is not just a question of theory. Courts already choose perspectives whenthey apply law to the Internet. They just don’t realize it. What the courts maysee as a minor skirmish in the “battle of analogies”130 is really a broaderproblem throughout Internet law. In areas ranging from criminal law to stateaction to copyright infringement, the clash between internal and external perspec-tives forces the courts to pick a perspective to reach a legal outcome. Thequestion is not whether courts should choose a perspective, for already they do,and perhaps must.

The real question, I think, is how courts should resolve the choice ofperspective once they recognize it. I think we need a way of thinking about howto choose whether to apply an internal or external perspective in a given case. Insome ways, this presents a new wrinkle to an old problem. Although lawyersoften think of “the facts” as distinct from “the law,” our way of arriving at thefacts can present important legal questions. The law of evidence provides ahelpful example. The field of evidence law presents an elaborate set of legalrules designed to govern how decisionmakers arrive at the facts.131 Evidencelaw recognizes that law and facts can be intertwined; the legal system does notrecognize a fact as a fact unless the law allows it to be recognized.

The problem of perspective calls for a similar recognition. It requires us tostop and question how we arrive at the facts of the Internet before we apply lawto it. Moreover, I think its resolution can harness principles that resemble thoseunderlying evidence law. Evidence law erects a screen to regulate what evi-dence a fact-finder can use to help build an accurate and reliable picture of thefacts.132 Similarly, we can develop a legal framework to resolve the problem of

cyberspace, we see matters afresh. To risk a metaphor from another technology, the Internetcan be an excellent lens for seeing other things. It is not, however, a particularly useful focalplane of legal analysis.

Sommer, supra note 122, at 1149.130. Mark S. Kende, The Supreme Court’s Approach to the First Amendment in Cyberspace: Free

Speech as Technology’s Hand-Maiden, 14 CONST. COMMENT. 465, 465 (1997).131. See ERIC D. GREEN & CHARLES R. NESSON, PROBLEMS, CASES AND MATERIALS ON EVIDENCE 25 (2d

ed. 1994) (“The function of the law of evidence is to specify . . . what types of information may beconsidered by the triers of fact in our law courts so that they may ascertain facts of importance to thedetermination of the dispute before the court[.]”).

132. See id.


perspective in concrete cases to help produce an accurate and reliable transla-tion of the facts from the physical world to the Internet.

While this may sound like a call for a grand theory of everything, in manyways the task is a very narrow one. The problem of perspective asks only howwe model the facts. As I suggested earlier,133 the problem of perspectivecoexists peacefully with broader struggles within cyberlaw that have occupiedboth cyberlaw proponents and skeptics. For example, we need not delve intowhether or how the Internet should transform individual legal doctrines.134

Perhaps the Internet should trigger major changes in First Amendment law, orcopyright law, or the state action doctrine. Perhaps it should not.135 Theproblem of perspective focuses on a much more humble and practical problem:modeling the facts so we can apply the law even as it currently exists to thefacts of the Internet. Regardless of what the law is or should become, theproblem of perspective remains.

This Part will explore the importance of the problem of perspective bysituating it within a broader set of theoretical premises in law, the socialsciences, and in theories of the Internet. My goal is to show how the problem ofperspective shares certain common themes with past technologies and scholar-ship, but also presents a significant new twist with important new practicalimplications. I will begin with technology, and show how the problem ofperspective reflects the universalization of themes that surfaced in very limitedform with prior technologies such as the telephone. I will next turn to the socialsciences, and show how the internal-external critique complements and extendsanalogous theories of systems advanced by social scientists and legal scholarssuch as H.L.A. Hart. I will conclude by showing how the problem of perspec-tive extends beyond past writings on the role of metaphor in Internet law.

A. PRELUDES TO THE PROBLEM OF PERSPECTIVE:AN EXAMPLE FROM THE TELEPHONE NETWORK

Because the problem of perspective derives from the existence of a virtualreality, we should expect to encounter the problem with technologies other thanthe Internet that also generate forms of virtual reality. This has two importantimplications, depending on whether we look forward or backward in time.Looking to the future, the problem of perspective should become more impor-tant as technology continues to advance and virtual realities become morewidespread and convincing. Conversely, looking back, preludes to the problemof perspective should appear in technologies that predate the Internet.

The telephone network provides the clearest example of a past technologythat presaged the problem of perspective. Perhaps the earliest case of competinginternal and external perspectives is the famous telephone wiretapping case of

133. See supra section I.C.134. See, e.g., Hardy, supra note 127, at 995.135. I consider myself enough of a cyberlaw skeptic to believe that it probably should not.


Olmstead v. United States,136 decided in 1928. In Olmstead, government agentswiretapped the telephone lines of a former police officer who operated amassive bootlegging operation in violation of the Prohibition laws.137 Theagents tapped the phone lines from a city street without entering onto anyprivate property. At trial and then on appeal, Olmstead argued that the wiretap-ping violated his Fourth Amendment rights.

Writing for a 5-4 majority, Chief Justice Taft rejected Olmstead’s argumentby applying what amounted to an external perspective of the telephone network.To Taft, the telephone network consisted of electrical lines that permitted itsusers to send communications out into the world.138 By using a telephone, Taftreasoned, Olmstead and his co-conspirators had opted to send their communica-tions out from the protected spaces of their houses and into the unprotectedspace of the public city street:

The reasonable view is that one who installs in his house a telephoneinstrument with connecting wires intends to project his voice to those quiteoutside, and that the wires beyond his house, and messages while passing overthem, are not within the protection of the Fourth Amendment. Here those whointercepted the projected voices were not in the house of either party to theconversation.139

The relevant question from Taft’s external perspective was how the networkworked, where the communications traveled, and where the agents were whenthey intercepted the communications. Because the telephone network workedby carrying the communications out into public spaces, using the phone was thefunctional equivalent of shouting out conversations where anyone could hearthem.140

In contrast, Justice Brandeis’s dissent offered a primarily internal account ofthe same event. To Brandeis, it was “immaterial where the physical connectionwith the telephone wires leading into the defendants’ premises was made.”141

Rather, the proper question was whether from a telephone user’s perspective,the wiretapping appeared as the equivalent of a search and seizure. Brandeis

136. 277 U.S. 438 (1928).137. Professor Murphy offers an excellent explanation of the Olmstead case and its context in his

book. See generally WALTER MURPHY, WIRETAPPING ON TRIAL (1965).138. See 277 U.S. at 465. Taft wrote:

By the invention of the telephone 50 years ago, and its application for the purpose ofextending communications, one can talk with another at a far distant place. The language ofthe [Fourth] amendment cannot be extended and expanded to include telephone wires,reaching to the whole world from the defendant’s house or office. The intervening wires arenot part of his house or office, any more than are the highways along which they are stretched.

Id.139. Id. at 466.140. See id.141. Id. at 479 (Brandeis, J., dissenting).


thought that it did: “Whenever a telephone line is tapped, the privacy of thepersons at both ends of the line is invaded, and all conversations between themupon any subject, and although proper, confidential, and privileged, may beoverheard.”142

Although Justice Brandeis did not predict the internal perspective of cyber-space, he came close: Brandeis seemingly understood the telephone network asa technological means of creating a private space for its users: a virtual“closet”143 where secrets could be “whispered.”144 The divide between Taft andBrandeis was not so much a difference between wooden and dynamic constitu-tional interpretation, as some commentators suggest,145 as it was a clash ofperspectives. Taft applied an external perspective of the telephone network, andBrandeis used an internal perspective.

It is tempting to view Olmstead as proof that the problem of perspective isreally nothing new. This is only partly correct. Olmstead reveals an earlyantecedent of the problem of perspective. However, it is a fairly isolated case.The books are not filled with cases applying law to “phonespace” in the sameway that we find cases applying law to cyberspace. This largely reflects thetelephone network’s narrow function; because the telephone merely transmitssound from one place to another, its ability to generate a virtual reality is fairlylimited. As a result, telephone cases raising an internal-external dynamic remainrare,146 and considered as a whole, they do not produce a recurring problem ofperspective. In contrast, the clash between the virtual and physical does seem toappear throughout Internet law. The advanced technology of the Internet hasuniversalized a problem that remained mostly latent in the early telephonenetwork. This may or may not make the problem of perspective truly “new.”Either way, the problem recurs often enough in Internet law that it prompts us toconfront it across a wide range of substantive areas.

B. BEYOND INTERNAL AND EXTERNAL PERSPECTIVES

IN LAW AND THE SOCIAL SCIENCES

Social scientists often use the terms “internal” and “external” to comparedifferent ways of analyzing systems such as religion and law.147 The internal

142. Id. at 475–76 (Brandeis, J., dissenting).143. Id. at 473 (Brandeis, J., dissenting).144. Id. (Brandeis, J., dissenting).145. See, e.g., Shirley M. Hufstedler, Invisible Searches for Intangible Things: Regulation of

Governmental Information Gathering, 127 U. PA. L. REV. 1483, 1495 (1979).146. Beyond Olmstead, we can also see the internal-external dynamic in Smith v. Maryland, 442

U.S. 735 (1979), discussed infra at subsection IV.C.2.147. See, e.g., Scott Brewer, Scientific Expert Testimony and Intellectual Due Process, 107 YALE

L. J. 1535, 1568–72 (1998) (summarizing scholarship on internal and external points of view); FrankCross, Political Science and the New Legal Realism: A Case of Unfortunate Interdisciplinary Igno-rance, 92 NW. U. L. REV. 251, 280–85 (1997) (same); Douglas Litowitz, Essay, Internal Versus ExternalPerspectives on Law: Toward Mediation, 26 FLA. ST. U. L. REV. 127, 131–35 (1998) (explaining the useof internal and external perspectives in the social sciences).


perspective offers the view of a participant in the system, who feels bound by itsrules; the external perspective adopts the view of a third-party observer whodoes not consider himself bound.148

Consider religion. When we discuss religion, we can talk as believers of acommon faith and try to understand our religious doctrines and our God.149

Alternatively, we can also approach religion as outsiders. We can talk about ournext-door neighbors and their unfamiliar religious practices and try to under-stand their beliefs and practices as reflections of history or culture.150 Theformer approach reflects an internal perspective of religion; the latter, anexternal perspective.151

The same construct has been applied to the legal system, most famously byH.L.A. Hart in The Concept of Law.152 According to Hart, an internal perspec-tive of the legal system accepts that we are bound by the rule of law, andindicates faith in the power and authority of legal reasoning and doctrine.153 Incontrast, those who view law from an external perspective see the rules asmerely window-dressing for other forces that generate “observable regularitiesof behavior”154 but have little additional significance. Hart uses the example ofa stop sign at an intersection. From an internal perspective, a driver sees a stopsign as the government’s order to stop his vehicle before he reaches theintersection. The stop sign has the force of law. In contrast, an external observerlooks at a stop sign and merely notes that most drivers stop their vehicles whenthey approach the sign. The external observer may recognize that driversperceive that they are bound by the sign, but does not find such perceptionsparticularly relevant.155

The problem of perspective in Internet law resembles these internal-externalcritiques, but it also features important differences. Like these critiques, theproblem of perspective compares the viewpoint of a participant in a system tothe viewpoint of a third-party observer.156 The “system” here is the Internetitself. The participant’s internal perspective adopts the user’s understanding ofthe virtual reality of cyberspace, and the observer’s external perspective adoptsa third party’s understanding of the physical network of the Internet.

However, the problem of perspective differs from past approaches in impor-

148. See Litowitz, supra note 147, at 133.149. See Stanley Ingber, Religion or Ideology: A Needed Clarification of the Religion Clauses, 41

STAN. L. REV. 233, 282 n.305 (1989) (“Those approaching religion from the internal viewpoint—thefaithful—will perceive their religious responsibilities not as an issue of individual choice and reasonbut of obligation imposed by a sacred or divine force.”).

150. See id. (“The external perspective may view religion as: a human creation, . . . a set of mythssurrounding a confused ideological claim, . . . or an opiate sedating an oppressed people.”).

151. See id; see also Litowitz, supra note 147, at 132.152. See H.L.A. HART, THE CONCEPT OF LAW 89–91 (2nd ed. 1994).153. See id. at 89.154. Id.155. See id. at 90.156. See Kerr, supra note 14, at 1290 n.7.


tant ways. The two perspectives of the Internet reflect distinct representations ofreality, rather than two ways to draw lessons from the same representation. Ithink this is different from H.L.A Hart’s critique. When Hart compares internaland external perspectives of law, he compares two lessons that can be drawnfrom a single picture. A stop sign can be understood either as an order from thesovereign or as an indicator that drivers will often stop their vehicles at aparticular place. Either way, however, a stop sign is still a stop sign. The onlyquestion is what significance we choose to assign to it.157

In the case of the Internet, however, the two perspectives follow two differentrepresentations of reality. The external perspective follows physical reality, andthe internal perspective follows virtual reality. So, for example, visiting awebsite can be either sending a request to a remote server that sends back textand pictures (physical reality), or traveling to a place (virtual reality). Theinternal viewer conceptualizes a different world than does an external viewer.Of course, a user may be aware of both perspectives simultaneously. Technically-savvy users who understand the technology are especially adept at following theexternal view along with the internal.158 But, nonetheless, the technology of theInternet generates a choice between two representations of reality: internalversus external, virtual versus real.159 A user may be aware of both realities atthe same time, but will choose to accept either the virtual world or the real onewhen trying to understand her experiences online.

C. BEYOND METAPHORS IN INTERNET LAW

A second intellectual cousin of the problem of perspective can be found inrecent scholarly writings on the role of metaphor in Internet law. In an impor-tant article published in 1995, Professor Michael Froomkin noted that the natureof Internet law “depend[s] critically on the legal metaphors” that are used todescribe the Internet and its functions.160 Froomkin’s article analyzed theconstitutionality of encryption regulations that the Clinton Administration pro-posed (but never enacted) in the early- to mid-1990s.161 Froomkin predicted

157. See HART, supra note 152, at 89.158. My own limited and unscientific survey suggests that technologically-savvy Internet users tend

to be more willing to embrace an external perspective, and computer novices tend to be drawn more tothe internal perspective. I think the best explanation for this trend is that novices first learn how to usecomputers and the Internet by developing an internal understanding of the Internet. In contrast,computer experts know how the Internet works; to them, the external perspective is simply reality,unchanged by whether a user happens to understand it.

159. The Matrix provides an extreme example. In the movie, discussed supra section I.A, theconnection to the network was total, and those who connected themselves to the network lost all contactor sense of external reality.

160. A. Michael Froomkin, The Metaphor is the Key: Cryptography, the Clipper Chip, and theConstitution, 143 U. PA. L. REV. 709, 718 (1995). For a similar approach, see O’Rourke, supra note 25,at 561.

161. Encryption is a process of encoding communications by scrambling them, usually followed bya process of decoding and thus descrambling them so they can be read. See generally SIMON SINGH, THE

CODE BOOK (1999).


that the degree of constitutional protection the courts afforded to the use ofencryption would hinge upon the metaphor the courts chose to describe it. Thecourts could analogize encryption regulation to the regulation of automo-biles,162 languages,163 safes,164 or houses,165 depending on the factual similari-ties and differences that the courts decided to emphasize. So, for example, if thecourts analogized encryption to a language, they would protect it from regula-tion because the First Amendment protects languages as speech.166 If the courtsanalogized encryption to an envelope around a message, however, they wouldbe more likely to uphold the regulation of encryption because the regulation ofenvelopes would not raise similar constitutional problems.167

The problem of perspective shares an intellectual thread with ProfessorFroomkin’s writings on the role of metaphor. Like Froomkin, I am concernedwith how we conceptualize new technologies when we apply law to them. Atthe same time, I think the problem of perspective differs significantly from thequestions considered by Froomkin. Froomkin takes the facts of the Internet as agiven and shows that its new facts can be compared to those of past technolo-gies in various ways. As Froomkin notes, this is an important point; theanalogies we draw can have legal consequences. However, this insight is notInternet-specific. As cyberlaw skeptics have rightly noted, we can always drawanalogies between new technologies and earlier ones.168 Recall Justice Cardo-zo’s famous opinion in McPherson v. Buick.169 McPherson considered anautomobile company’s tort liability for selling a car with defective wheelsmanufactured by another company.170 As explained by Cardozo, the questionboiled down to the hunt for the best analogy: Was an automobile like a stagecoach, in which case the manufacturer would not be held liable? Or was it morelike a railroad, in which case it would?171

Even Professor Froomkin’s specific inquiry into choosing metaphors forencrypted communications turns out to have predated the Internet by at least acentury.172 Encryption has existed for hundreds of years and was used to encodetelegraph communications in the nineteenth century and letters in the eighteenth

162. See Froomkin, supra note 160, at 863–65.163. See id. at 865–70.164. See id. at 871–74.165. See id. at 874–79.166. See id. at 884 (“If the courts treat a ciphertext as if it had been written in a foreign language, it

will trigger a First Amendment analysis that will result in giving cryptography more protection.”).167. See id. (“If encryption is considered no more than an outer envelope in a message transmission

system . . . it is likely to receive the lowest level of protection.”).168. See Sommer, supra note 122, at 1148.169. 217 N.Y. 382 (N.Y. 1916).170. Id. at 391.171. Justice Cardozo preferred the latter analogy. The defective wheels, he wrote, were “as much a

thing of danger as a defective engine for a railroad. . . . Precedents drawn from the days of travel bystage coach do not fit the conditions of travel to-day.” Id. at 391.

172. See Kerr, supra note 57, at 526 (noting the use of encryption as far back as the ancient Greeks).


century.173 Then, as now, courts needed to select different metaphors to describeencrypted communications. So while Froomkin predicts that courts may someday have to grapple with whether encrypted communications are like communi-cations in a foreign language, he overlooks the fact that century-old casesstruggle with this precise question.174

In contrast, computer networks and the Internet provide the first technologyin which the problem of perspective predominates. Although the telephonepresaged similar issues (as in Olmstead), the Internet is the first technology ineveryday use that supports a virtual reality for its users sufficiently realistic towarrant its own spatial metaphor, cyberspace.175 For the first time, the virtualmetaphor is competing with the physical world, creating choices between whichview to adopt when applying the law to the Internet.176

173. See id. at 526–29 (describing popular uses of encryption in the eighteenth and nineteenthcenturies).

174. One fascinating example is Western Union Telegraph Co. v. Olivarri, 110 S.W. 930 (Tex. Civ.App. 1908), aff’d 135 S.W. 1158 (Tex. 1911). In Olivarri, a woman in Texas sent a telegram in Spanishto her husband in Mexico informing him that her newborn children were ill and likely would die. See110 S.W. at 931. Western Union failed to deliver the message, and the woman sued Western Union forher pain and suffering when her husband did not arrive to take care of her during her children’s deaths.See id. The established rule of law was that the telegraph company could be liable for failure to delivera telegram correctly when the importance of its message was clear on its face. See id. at 932. Under thisrule, telegraph companies were not liable for the consequences of misdelivered encrypted communica-tions: By their nature, the importance of such communications was not clear to the telegraph company.See Einbinder v. Western Union Tele. Co., 30 S.E.2d 765, 773 (S.C. 1944).

In its defense, Western Union argued that messages in a foreign language were like encryptedmessages, or, as they termed it, “cipher messages.” See Olivarri, 110 S.W. at 932. The Texas Court ofCivil Appeals rejected the metaphor:

We do not think a message in a foreign tongue for delivery in a country where that tongue iswritten and spoken can be placed in the same category as a cipher message. That kind ofmessage is sent for the purpose of concealing from the telegraph company, as well as all otherparties, except the person to whom it is sent, the purport of the message. The telegraphcompany, not being in possession of the key to their meaning, cannot possibly understandsuch telegrams, and is under no obligation to make any inquiries in regard to them. In factdisclosure of their meaning would defeat their very object in sending them. The telegram,however, to a person in a foreign country in the language of that country, is not intended toconceal.

Id.175. See, e.g., Peter H. Lewis, In Search of a Meaningful Relationship with Computers, N.Y. TIMES,

May 27, 1990, at E5. (“One intriguing area of metaphor exploration is in the field of cyberspace, wherethe user enters a ‘virtual reality’ that exists as pure computer data.”).

176. In a sense, Froomkin’s analysis echoes H.L.A. Hart’s: Both focus on ways of understandingwhat we see, rather than questioning how we model what we see. The problem of perspective insteadgrapples with what we see when we look at the Internet: its physical reality, or its virtual reality. Thechoice of perspective can influence whether a metaphor appears convincing, of course. For example,describing encryption as a safe that locks a communication offers an internal metaphor; a user sees thebrowser “lock” and “unlock” the file, even though the encryption does not actually lock anything in theexternal physical world. The choice of perspective asks an antecedent question to the search for theright metaphor: Should we search for metaphors from the perspective of an Internet user, or from theperspective of a third-party observer?


IV. A NORMATIVE FRAMEWORK FOR RESOLVING THE PROBLEM OF PERSPECTIVE

Because perspective exerts a significant influence on Internet law, two obvi-ous questions arise. Are both perspectives equally legitimate in every case? Andif not, how do we choose whether to apply an internal or external perspective ina given case? This Part attempts to answer these questions. My goal is toconvince the reader that while the problem of perspective does add a new layerof inquiry when we apply law to the Internet, it is not an irresolvable one. Theproblem of perspective does not plunge us into hopeless indeterminacy; it is aproblem, but not one without answers. Once we agree on our goals whenapplying law to the Internet, we can set up a framework for resolving theproblem of perspective that can help us determine the best answer in mostcases.

This Part will propose such a normative framework, based on the traditionaljudicial goal of fidelity to existing law. The normative framework offers twoways of selecting perspectives. The first approach tasks the decisionmaker withstudying the applicable law for clues as to whether the doctrine reflects aninternal or external approach and then applying the perspective the doctrineexplicitly or implicitly adopts. The second approach provides a backup whenthe first approach does not yield an answer, and suggests that we should adoptthe perspective of the party that the law seeks to regulate. Both approaches seekto resolve the problem of perspective by reference to the substance of functionof existing law; they view the problem of perspective as a problem of fidelityand seek solutions that adhere closely to existing law.

An important caveat is in order before I begin. The framework I present willnot be the final word on the problem of perspective. First, some will object tomy goal of fidelity to law; especially in academic circles, it will strike some asnarrow and (gasp!) lawyerly. Even those who accept my goal may find myapproaches not entirely satisfying. Indeed, while I think the first approach isentirely sound, I am not entirely sure of the second—or even that it is entirelydistinct from the first. Perhaps a third, yet-undiscovered method may also work.In light of my own uncertainty, my limited aim will be to explain a fewimportant features that could guide an analytical framework for choosingperspectives, rather than present a foolproof recipe for churning out the “right”perspective in a given case. I hope to offer a starting point and begin a dialogue,not resolve the problem of perspective once and for all.

A. THE NATURE OF THE CHOICE AND THE PROBLEM OF INDETERMINACY

It may be tempting to see the problem of perspective as the final nail in thedeterminacy coffin.177 If we can reach one outcome from the internal perspec-

177. See, e.g., Steven L. Winter, Indeterminacy and Incommensurability in Constitutional Law, 78CAL. L. REV. 1441, 1445 (1990); see also Anthony D’Amato, Pragmatic Indeterminacy, 85 NW. U. L.REV. 148 passim (1990). For a critique, see Richard Epstein, Some Doubts on Constitutional Indetermi-nacy, 19 HARV. J. L. & PUB. POL’Y 363, 364 (1995).


tive and the opposite outcome from an external perspective, then isn’t the lawhopelessly indeterminate? Won’t courts simply apply the perspective they wantin order to reach the result they prefer in a given case?178 Is the search for the“right” perspective a pipe dream, as the two perspectives will prove equallyplausible in every case?

I think the answer to these questions is “no.” The fact that we can alwayspresent an internal and external version of the facts does not mean that thosetwo versions are equally viable, any more than the fact that there are two sidesto every argument means that we can never choose between them. Judgmentscan be made.

The trick, I think, is that first we need to confront what our goals are when weapply law to the Internet. Once we agree on our goals, we can use them todevelop a framework for choosing between perspectives. Consider a simpleexample. Let’s say our goal is to foster a substantive vision of what cyberspaceshould be.179 For example, we may want cyberspace to be “free” and “open,” sothat Internet users perceive few limits on the range of experiences they can haveonline. If this is our goal, then solving the problem of perspective can becomequite easy. Simply pick the perspective that in the context of that case will helpfoster the substantive vision, in this case a free cyberspace.180 If designatingAOL a state actor will make cyberspace more free by limiting AOL’s right tocensor speech, then we can take an internal perspective that will help producethat result. I do not mean to endorse this approach; to me, choosing a perspec-tive based on a substantive vision of cyberspace seems perilously close to justdoing whatever you like. But as an illustration, the point sticks. Once we selectour goals, we can use those goals to frame a solution.

Among legal theorists, this may not seem a major step forward. The difficultyis that what our goals should be when we apply law is (to state it mildly)contested ground. Different theorists offer different goals—to borrow againfrom H.L.A. Hart, different rules of recognition181—that they argue shouldgovern. I do not pretend to have a way to resolve these differences. Nor do Ihave my own metatheory to throw on to the already-crowded pile. Rather, Ihope to sidestep this jurisprudential thicket and instead focus on a morepractical question, which I find more important and more interesting. In the real

178. See, e.g., Christopher F. Edley, Jr., The Governance Crisis, Legal Theory, and PoliticalIdeology, 1991 DUKE. L.J. 561, 580 (1991) (quoting Charles Evans Hughes’s comment that “[w]e areunder a Constitution, but the Constitution is what the judges say it is.”).

179. Professor Lessig’s writings at times suggest such an approach. See, e.g., LESSIG, supra note 20,at 223 (“In cases of simple translation—where there are no latent ambiguities and our traditions seem tospeak clearly—judges should firmly advance arguments that seek to preserve original values of libertyin a new context. In these cases there is an important space for activism. Judges should identify ourvalues and defend them.”).

180. See id.181. See HART, supra note 152, at 100–10. In Hart’s jurisprudence, a rule of recognition is the

foundational rule that justifies the remaining rules in applying the law, but that itself is based primarilyon widely accepted and often unstated social practice. See id.


world of courts and litigants, where lawyers and judges routinely apply the lawto new facts, how should they resolve the problem of perspective? Embracingthe conventional goal of applying the law to new facts faithfully and accurately,how can we best try to resolve the problem of perspective?

The remainder of this Part proposes a normative framework for resolving theproblem of perspective based on fidelity to law.182 It starts with the importanceof doctrine, rooted in fidelity to text, and then turns to the benefits of trackingthe perspective of the individual the law seeks to regulate, rooted in fidelity tofunction.

B. A FIRST APPROACH: RECOGNIZING THE POWER OF DOCTRINE

The most important insight that helps resolve the problem of perspective isthat if our goal is fidelity to law, that law will often signal the adoption of aninternal or external approach. When mined for clues, traditional legal sourceswill often indicate whether the law is more closely attuned to external orinternal concerns. As a result, doctrine itself can often help resolve the problemof perspective. Once we are aware of the problem of perspective, the existinglaw can lead us to either an internal or external approach depending on theparticular case. In such cases, choosing a perspective is akin to the task ofstatutory interpretation: Just as we go no further when the text of a statuteestablishes its meaning,183 we should look no further when the doctrine estab-lishes the proper perspective. When the applicable law derives from a statute,these two rules converge into one.

To show how this might work in practice, I will examine three examples ofhow doctrine itself can resolve the problem of perspective. I will start with afairly straightforward hypothetical, in which the statutory law reflects a choiceof perspective, move next to a case in which statutory law implicitly adopts aperspective, and conclude with a more complex example, in which the view-

182. I recognize that my use of the word “fidelity” may bring to mind Professor Lessig’s theory oftranslation. See, e.g., LESSIG, supra note 20, at 111–21. Lessig’s theory of translation is a metatheory forapplying constitutional protections to the Internet. Although Lessig does not offer many examples ofhow translation would actually work, it appears from Lessig’s writings that a judge attempting totranslate constitutional protections should attempt to abstract constitutional text into the policy concernthat represents the essence of the protection, and then create a new rule that enforces that policyeffectively. See id. at 118. So for example, Lessig understands the Fourth Amendment as a protector ofprivacy, and suggests that a judge should apply the Fourth Amendment to the Internet by devising rulesthat protect privacy, along the lines of Justice Brandeis’s dissent in Olmstead. See id. Professor Lessigargues that this approach is the best way of maintaining fidelity to the Constitution because it remainsloyal to the policy concerns that motivated the Framers. See id. at 115.

The fidelity that I speak of in this section is a much more ordinary type of fidelity. My concern is notfidelity to a conception of policy concerns that motivated the Framers of the Constitution, but merelyfidelity to existing rules of law—regardless of whether they are constitutional, statutory, common law,or regulatory in origin.

183. See, e.g., Community For Creative Non-Violence v. Reid, 490 U.S. 730, 739 (1989) (“Thestarting point for our interpretation of a statute is always its language.”); Oliver Wendell Holmes, TheTheory of Legal Interpretation, 12 HARV. L. REV. 417, 419 (1899) (“We do not inquire what thelegislature meant; we ask only what the statute means.”).


point must be teased out of a body of case law. In all three cases, a careful studyof the applicable law yields fairly clear answers to the question of whichperspective should apply. Although we can imagine both internal and externalapproaches in each case, the law itself points to which perspective is correct.

1. When the Legislature Has Expressly Chosen a Perspective: Variations onKammersell

The choice of perspective presents the least difficulty when the applicablelaw reveals an express decision to adopt a particular perspective. ConsiderUnited States v. Kammersell,184 the case involving an instant message bombthreat that traveled from Utah to Virginia, and then back to Utah. The legal issuewas whether Kammersell sent an interstate threat. From Kammersell’s internalperspective, he had merely sent a threat within Utah; from the government’sexternal perspective, he had sent the threat from Utah to Virginia, and then backto Utah. Which is correct? The answer is up to Congress: Congress can write athreat statute that adopts an internal perspective, or it can write the same statutefrom an external perspective.185 Our task as lawyers and judges should properlybe to determine whether Congress has written the statute from an internal orexternal view and to apply the perspective that Congress has chosen.

We can see this by hypothesizing two ways Congress could have written thefederal threat statute. First, imagine that the statute prohibited “knowinglysending a threat, if the communication or signal that carries or constitutes thethreat crosses state lines in the course of delivery.” This language indicates aCongressional decision to adopt an external perspective; the doctrine focuses onthe external path of the signal across the network. While it would be possible toapproach this test from an internal perspective, it would seem highly artificial,as the text expressly delinks the sender’s mens rea from whether the threat crossstate lines. The text requires us to focus on the external question of the signal’spath, not the internal question of the sender’s perspective.186

Next imagine that Congress amends the statute, and that this time insteadchooses to make it a crime “to send a threat that the sender knows or believeswill travel across state lines.” Assuming that the sender will be online, this

184. 196 F.3d 1137 (10th Cir. 1999). Kammersall is discussed supra subsection II.C.1.185. Granted, interesting questions of Congressional power to enact such a statute would arise in

such a case. Congress probably lacks the authority to enact a general threat statute. See United States v.Lopez, 514 U.S. 549, 566 (1995) (holding that Congress exceeded its Commerce Clause powers inenacting the Gun-Free School Zones Act). However, early cases that construed the Commerce Clausebroadly to allow Congress to regulate the telephone system would presumably give Congress broadauthority to enact an Internet threat statute, even without an interstate communication requirement. See,e.g., Weiss v. United States, 308 U.S. 321, 327 (1939) (“Congress has power, when necessary for theprotection of interstate commerce, to regulate intrastate transactions.”).

186. I think this explains the Kammersell decision. See 196 F.3d at 1139. The current text of 18U.S.C. § 875(c) states that “[w]hoever transmits in interstate or foreign commerce any communicationcontaining any threat to kidnap any person or any threat to injure the person of another, shall be finedunder this title or imprisoned not more than five years, or both.” This is an external standard; it looks towhether a communication containing a threat was transmitted in interstate or foreign commerce.


amendment switches us from an external to an internal perspective: Now whatmatters is the state of mind of the Internet user—what the sender knows orbelieves—rather than the technical details of how the network handled thecommunication. By requiring that the sender know or believe that the communi-cation will travel across state lines, the latter version of the text signals aCongressional focus on an internal perspective. The user’s own state of mindnow matters, triggering an internal analysis for internal Internet users.187

It is unlikely that Congress considered whether it wanted courts to follow aninternal or external perspective when it first enacted the interstate threat statutein 1934.188 However, the language it selected then may nonetheless point to aparticular perspective today. More importantly, Congress can draft future stat-utes with the problem of perspective in mind. By making the choice ofperspective itself, Congress can simplify the courts’ task dramatically andensure that the courts follow rather than defeat the legislative intent.

2. When the Legislature Has Implicitly Chosen a Perspective: CopyrightInfringement

A similar analysis applies when the doctrine does not expressly choose aperspective but its structure implicitly opts for one perspective over the other.Consider the copyright dispute between the major record companies andMP3.com.189 Recall that MP3.com started a service that provided Internet userswith music over the Internet. From an internal perspective, the service merelyprovided its users with a virtual storage locker of their CDs. From an externalperspective, however, MP3.com had copied copyrighted materials twice: first bycopying the music onto MP3.com’s servers, and then second by copying themusic from the servers to send off to the requesting Internet users. The questionis which perspective should we choose: the user’s or a third-party observer’s?

The answer can be found in the structure of the Copyright Act,190 and inparticular the computer-specific language of 17 U.S.C. § 117. As amended in1980, § 117 carves out a special exception for copies made by computers as anessential step toward using computer programs.191 The section provides that,notwithstanding the general rule of copyright liability,

it is not an infringement for the owner of a copy of a computer program tomake or authorize the making of another copy or adaptation of that computer

187. Importantly, this does not mean that a subjective standard is always internal, or that anobjective standard is external. A subjective standard will follow the perspective of the person whosesubjective mental state matters: When that person is online, the internal perspective will govern,whereas when that person is offline, the external perspective will apply. I discuss this approach later,infra Section IV.C.

188. See Pub. L. C. 300, 48 Stat. 781 (1934).189. See supra section II.D; UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349 (S.D.N.Y.

2000).190. See 17 U.S.C. § 101–1332 (2001).191. See Micro-Sparc, Inc. v. Amtype Corp., 592 F. Supp. 33, 34–35 (D. Mass. 1984).


program provided . . . [that] such a new copy or adaptation is created as anessential step in the utilization of the computer program in conjunction with amachine and that it is used in no other manner.192

Congress enacted this exception because using computer software requiresmaking a copy; any time a computer runs software, the computer must firstcopy the program from an external source into its memory.193 The rationale of§ 117 is that without this express exception, merely loading a copyrighted fileinto a computer’s memory would constitute a violation of the Copyright Act.194

The importance of § 117 lies in its strikingly external approach to copyrightlaw. Section 117 appears unnecessary from an internal perspective; a user doesnot perceive that the operating system has made a copy when it loads aprogram, just as a user of MP3.com does not perceive that MP3.com has made acopy of his music before sending it to him. As a result, we can read theenactment of § 117 as a signal that Congress has approached copyright infringe-ment from an external perspective. To avoid reducing § 117 to a nullity, and toremain faithful to Congress’s design, we should defer to this understanding andapproach questions of infringement in cyberspace from an external perspectiveas well. This method roughly tracks analogous rules of statutory interpretation.Just as we should avoid interpreting a statute in a way that denies effect toparticular statutory language,195 we can avoid choosing a perspective that hasthe effect of nullifying statutory text, in this case 17 U.S.C. § 117.

3. Teasing Perspective from Case Decisions: The State Action Doctrine

Using doctrinal signals to guide the choice of perspective proves morecomplicated outside the statutory realm, but the same basic insights apply.When the legal doctrine derives from the common law or the Constitution, wecan study it for clues about whether it appears more closely attuned to externalor internal concerns (or, if a bit of both, whether it seems more attuned to oneperspective than another). Not all areas will reveal a particular orientation.However, many areas will contain important clues that facilitate a decisionabout which perspective to adopt in a given case.

Consider the state action doctrine of constitutional law. As we saw in sectionII.B, private manufacturers and operators of Internet hardware and software canbe considered state actors from an internal perspective, but not from an externalperspective. How do we choose? We choose by navigating through the SupremeCourt’s cases that raised analogous conceptual issues. Litigants have often

192. See 17 U.S.C. § 117 (2000).193. See Micro-Sparc, 592 F. Supp. 2d at 35 (“The legislative history of § 117 indicates to us that

subsection (1) was enacted simply to permit the rightful possessor of a program to input and use it.”).194. See id. (citing FINAL REPORT OF THE NATIONAL COMMISSION ON NEW TECHNOLOGICAL USES OF

COPYRIGHTED WORKS 31 (1978)).195. See Kungys v. United States 485 U.S. 759, 778 (1988) (noting that “the cardinal rule of

statutory interpretation [is] that no provision should be construed to be entirely redundant”).


claimed that private action assumed the character of state action because fromthe individual’s perspective the private actors had assumed government func-tions. Most famously, the Supreme Court held in Marsh v. Alabama196 that theoperators of the “company town” of Chickasaw, Alabama could be state actors;although the operators were technically private employees of the Gulf Shipbuild-ing Company, they had assumed the role of the town’s governing body andtherefore could be treated as state actors for constitutional purposes.197

Marsh does not directly raise the problem of perspective. Like H.L.A. Hart’scritique, the Marsh case asks us to choose an interpretation of what we see inthe physical world, rather than choose between two different representations ofwhat we see. At the same time, Marsh resonates with the internal-externalcritique. Both Marsh and the problem of perspective consider whether andwhen the law should credit a participant’s view of the world around him. Inparticular, Marsh suggests that there may be room for an internal perspective inInternet law; an ISP or software company may “seem” like the government toan Internet user like the Gulf Shipbuilding Company “seemed” like the govern-ment to the residents of Chickasaw, Alabama.198

The remainder of the Court’s state action jurisprudence suggests, however,that an internal approach could apply in only extremely limited circumstances.While Marsh remains good law, the Supreme Court generally has rejected theMarsh approach when a private entity “seemed” like the government only tospecific individuals encountering the entity in the context of a specific relation-ship. For example, the Court has rejected the argument that the NationalCollegiate Athletic Association is a state actor because it acts as the governmentto college athletes and coaches;199 that the owner of a private shopping mall is astate actor because it acts as the government to its customers;200 that insurancecompanies can be state actors because they act as the government to employeesseeking worker’s compensation benefits;201 that a private utility company is astate actor because it acts as the government to its customers;202 and thatadministrators of a private school can be private actors because they act as the

196. 326 U.S. 501 (1946).197. Id. at 507–08.198. Unsurprisingly, the plaintiffs in Cyber Promotions v. AOL relied heavily upon Marsh in their

briefs. See 948 F. Supp. 436, 442 (E.D. Pa. 1996) (“Cyber therefore contends that AOL’s Internet e-mailaccessway is similar to the company town in [Marsh] which the Supreme Court found performed apublic function and therefore was a state actor.”).

199. See National Collegiate Athletic Ass’n v. Tarkanian, 488 U.S. 179, 195 (1988) (holding that theNCAA is not a state actor).

200. See Hudgens v. NLRB, 424 U.S. 507, 518–20 (1976) (holding that a shopping center is not astate actor), overruling Food Employees v. Logan Valley Plaza, 391 U.S. 308 (1968).

201. See Am. Mfrs. Mutual Ins. Co. v. Sullivan, 526 U.S. 40, 58 (1999) (holding that insurancecompanies did not act as state actors when they denied insurance benefits to claimants).

202. See Jackson v. Metro. Edison Co., 419 U.S. 345, 358–59 (1974) (holding that a utility companyis not a state actor).


government to its students.203 The Court has effectively adopted an externalviewpoint in these cases, describing the private entities based on their appear-ance to outsiders rather than the internal perceptions of their users and custom-ers.

These cases suggest that the courts should approach most (if not all) ques-tions of state action in cyberspace from an external perspective.204 The claimthat “within my world, this entity is like the government to me” is not enough totrigger state action, regardless of whether that “world” is the world of a collegeathlete, shopping mall visitor, student, or Internet user. Although Marsh leavesopen the possibility that an internal approach may be appropriate in some cases,the remainder of state action case law indicates that the norm when applying thestate action doctrine in cyberspace should be an external perspective.

C. A SECOND APPROACH: FOLLOWING THE PERSPECTIVE OF THE PERSON THE LAW

SEEKS TO REGULATE

Although doctrine guides the choice of perspective in most cases, it mayprove unilluminating in some. Doctrine alone will not always resolve theproblem of perspective. When it does not, a second strategy may prove helpful:applying the perspective of the individual whom the law seeks to regulate. Assuggested earlier, I am more cautious about this second approach than the first.Nonetheless, I think it provides a potentially useful way to approach theproblem of perspective when a close reading of doctrine proves insufficient.

The core theory behind applying the perspective of the regulated individual isthat legal rules set up regulatory schemes designed to influence the behavior oftargeted individuals.205 For example, traffic laws seek to regulate drivers;contract law seeks to regulate contracting parties. Because these laws attempt toregulate identifiable groups, we can best translate the function of our regulatoryschemes from the physical world to the Internet by following the perspective ofthe individual the law seeks to regulate as that individual interacts with theInternet. I have in mind an objective test, not a subjective one. The question isnot whether a particular individual subjectively understood the Internet in aparticular way, but rather whether his conduct was online or offline.206 Whenthe law seeks to regulate offline conduct, we can apply an external perspective;when it seeks to regulate online conduct, we can apply an internal perspective.

203. See Rendell-Baker v. Kohn, 457 U.S. 830, 843 (1982) (holding that a private school is not astate actor).

204. Accord Berman, supra note 67, at 1267 (“The state action doctrine . . . poses a significantchallenge to those who see private regulatory power as a threat to individual rights and public discourseonline.”).

205. See HART, supra note 152, at 10–20.206. This follows from the regulatory function that the law seeks to achieve. The law seeks to

regulate all of those in the party’s position, regardless of the subjective perspective that certainregulated individuals may adopt. Accordingly, it looks objectively to whether the regulated individualinteracts with the network as a user or a third-party.


The law’s perspective will track the perspective of the individual the law seeksto regulate.

As with the first method, I see this approach as a method rooted in fidelity toexisting law; it seeks to apply law to the Internet by transferring the law’sregulatory scheme as faithfully and completely as possible from the physicalworld to the Internet. It sees that the law imposes rules on regulated actors bytelling them what they can and cannot do, and ensures that the rules that theseactors encounter online are the same as those they encounter offline. Rather thanhold these actors to a foreign perspective they do not encounter and may beunable to predict,207 we can break a doctrinal tie between perspectives byfollowing the perspective that regulated actors encounter when they interactwith the Internet.

In a sense, this second approach is a more abstract version of the first one.Both methods are rooted in doctrine; in the first approach, we look to doctrine tosee whether the doctrine tracks internal or external concerns, and in the secondwe look to doctrine to see who the law regulates. The latter approach can helpwhen the former does not, however, because clues for whether doctrine tracksinternal or external concerns tend to be explicit, and such clues may or may notexist. In contrast, the law will usually (if not always) regulate a particular party,offering a second doctrinal hook when the first one fails.

1. An Example: The Fourth Amendment

The Fourth Amendment provides a useful example. Assume for now208 thatFourth Amendment doctrine does not necessarily resolve whether to apply aninternal or external approach. The question becomes, who does the FourthAmendment try to regulate? If we can identify who the Fourth Amendmentregulates, we can pick a perspective by following the viewpoint of that party asthey interact with the network.

207. Wiretapping provides a good example. Imagine a Fourth Amendment rule that the police canwiretap a wire carrying Internet traffic without a warrant only when the users of that line do not have areasonable expectation of privacy in their communications sent over that line as evaluated from theusers’ internal perspective. See 18 U.S.C. § 2511(2)(i) (2000) (creating a similar rule by statute innarrow circumstances). Under this rule, a police officer approaching the line externally cannot tap theline until he is satisfied that the traffic flowing through the line is not part of a communication that whenreconstructed into a user’s internal world would be protected under the Fourth Amendment. Thedifficulty is that the external police officer has absolutely no idea of whether a particular communica-tion would be protected by an internal Fourth Amendment rule until after he has tapped the line. Hewould need to tap the line first and reconstruct the communications later. Even then it may beimpossible to evaluate the effect of the tap on the user’s privacy rights, as the police officer would haveno way of knowing the circumstances of the sender or receiver of the communication. If the Internetuser is at home, he presumably would have a reasonable expectation of privacy; if he is in prison, hewould not. Holding a police officer to this internal standard is unworkable from the officer’s externalstandpoint; he would have no idea of the effect that his external conduct would have on internalconcerns, and therefore no idea of when he can tap a line. See generally Kerr, supra note 14, at1298–1300.

208. I will return to this point later; see infra section IV.D.


According to the Supreme Court, “Fourth Amendment doctrine . . . is primar-ily intended to regulate the police in their day-to-day activities.”209 The rules setup a regulatory scheme that limits how the government can collect evidence ofcrime.210 Although the underlying purpose of the Fourth Amendment is toprotect privacy,211 Fourth Amendment doctrine protects privacy by regulatinghow the government investigates crime. In developing that regulatory scheme,the Court strives to create readily administrable rules212 that the police canfollow “on the spur (and in the heat) of the moment.”213 Viewing FourthAmendment law as a scheme for regulating the police suggests that we may tryto choose a perspective in Fourth Amendment cases by following the perspec-tive of the police. When the police officer is offline, apply an external perspec-tive; when the police officer is online, switch to an internal perspective.

Although cases applying the Fourth Amendment to the Internet remain quitesparse, the few decided cases suggest that the courts have done just that. Forexample, in United States v. Charbonneau,214 a Florida police officer workingwith an FBI undercover task force went online posing as a pedophile named“Mikey1L.”215 Mikey1L visited AOL chat rooms devoted to child pornographyand recorded everything he observed.216 At one point when Mikey1L waslogged in to one of the chat rooms, another chatroom visitor named “Charbyq”sent an e-mail to everyone else in the chat room that included an attachmentcontaining child pornography. The FBI then traced the “Charbyq” account toKenneth Charbonneau, a resident of Columbus, Ohio, and charged him withdistributing child pornography.217 Charbonneau responded with a motion tosuppress, in which he claimed that the undercover police officer had violated his“reasonable expectation of privacy” by recording everything in the chat roomand receiving Charbonneau’s e-mail under false pretenses.218

The court adopted the officer’s internal perspective and rejected Charbon-neau’s argument. The court analogized the chat room to a physical room, andthe online police officer to an undercover agent located “in” the chat room.From this virtual perspective, the facts closely matched those in Hoffa v. UnitedStates,219 a Supreme Court case in which mobster Jimmy Hoffa had spoken of

209. New York v. Belton, 453 U.S. 454, 458 (1981) (quoting Wayne LaFave, “Case-By-CaseAdjudication” Versus “Standardized Procedures”: The Robinson Dilemma, 1974 S. CT. REV. 127, 142).

210. See William J. Stuntz., Reply, 93 MICH. L. REV. 1102, 1103 (1995) (“It is common ground inFourth Amendment law and literature . . . that its primary purpose should be to regulate what policeofficers can see and hear.”).

211. See Warden v. Hayden, 387 U.S. 295, 304 (1967) (“[T]he principal object of the FourthAmendment is the protection of privacy.”).

212. See Atwater v. City of Lago Vista, 532 U.S. 318, 347 (2001).213. Id.214. 979 F. Supp. 1177 (S.D. Ohio 1997).215. Id. at 1179.216. Id.217. Id.218. Id. at 1183.219. 385 U.S. 293 (1966).


his crimes amidst a colleague who turned out to be a government informant.220

The Court held that the Fourth Amendment did not protect Hoffa against hismisplaced confidences; Hoffa had assumed the risk that the friends amongwhom he spoke would turn against him.221 To the Charbonneau court, anundercover agent logged in to a chat room required a straightforward applica-tion222 of Hoffa:

All of the evidence gathered by the FBI from the chat rooms resulted from thepresence of undercover agents in the rooms. Clearly, when Defendant en-gaged in chat room conversations, he ran the risk of speaking to an under-cover agent. Furthermore, Defendant could not have a reasonable expectationof privacy in the chat rooms. Accordingly, the e-mail sent by Defendant toothers in a “chat room” is not afforded any semblance of privacy; thegovernment may present the evidence at trial.223

Note that the court did not approach the chat room externally; it did notdescribe the “chat room” as a computer program hosted by remote computerservers that manage the communications among users logged in to the net-work.224 Instead, the court applied an internal perspective and understood thechat room as a virtual room.225

Courts have approached the Fourth Amendment quite differently when lawenforcement agents interact with computer networks as outsiders. Bohach v.City of Reno226 provides a helpful illustration. In this case, investigators fromthe Reno, Nevada Police Department conducted an internal affairs investigationinto possible misconduct by two police officers, John Bohach and Jon Catalano.The investigators sought copies of incriminating alphanumeric pages that hadbeen sent between the two officers using the police department computernetwork. The department computer network included a special paging programcalled “Alphapage,” which worked by storing copies of pages on a centralserver and then broadcasting the page over radio frequencies.227

220. Id. at 298.221. Id. at 310.222. See Charbonneau, 979 F. Supp. at 1184.223. Id. at 1185.224. See MULLER, supra note 12, at 31–37 (explaining how chat rooms work).225. Courts have also applied an internal perspective to law enforcement searches of computer files

stored on a hard drive. The courts have analogized the files stored on a hard drive to individual closedcontainers, and have copied the Fourth Amendment rules that apply to accessing containers to the caseof accessing files. See, e.g., United States v. Runyan, 275 F.3d 449, 464 (5th Cir. 2001); United States v.Barth, 26 F. Supp. 2d 929, 936–37 (W.D. Tex. 1998); United States v. Reyes, 922 F. Supp. 818, 832–33(S.D.N.Y. 1996); United States v. Lynch, 908 F. Supp. 284, 287 (D.V.I. 1995); United States v. Chan,830 F. Supp. 531, 535 (N.D. Cal. 1993); United States v. Blas, No. 90-CR-162, 1990 WL 265179, at*21 (E.D. Wis. Dec. 4, 1990). Rather than focusing on the external picture of how the magnets collectdata from across the hard drive and piped to computer memory, courts have adopted the internal pictureof opened and closed containers.

226. 932 F. Supp. 1232 (D. Nev. 1996).227. Id. at 1234.


The internal affairs investigators obtained copies of the pages between Bo-hach and Catalano by accessing the stored files that were retained on thenetwork server. Bohach and Catalano sued, claiming that the investigators hadviolated the Fourth Amendment by accessing their private communicationswithout a warrant.228 The court applied an external perspective and rejected theofficers’ Fourth Amendment claims. The Alphapage system automatically re-corded and stored the contents of all pages sent over the computer network on acentralized server, the court noted, “not because anyone is ‘tapping’ the system,but simply because that’s how the system works.”229 To the investigators, thepolice computer network was merely a repository for information that theycould use to collect evidence for their case. The court matched their externalperspective of the network with an external application of the Fourth Amend-ment.

At first blush, Charbonneau and Bohach may appear inconsistent with eachother. Bohach applied an external perspective, Charbonneau an internal perspec-tive. However, I think the two are quite consistent, in that each applied theperspective of the individual the law seeks to regulate—here, law enforcementofficers searching for evidence of crimes—and applied the Fourth Amendment’sregulatory framework to the version of the world that the officers encounteredas they interacted with the network. In Charbonneau, the police officer inter-acted with the Internet as a participant, logging on to the Internet and entering achat room. In Bohach, the investigators approached the network as externalobservers, viewing the network as a collection of physical boxes and wires thathad stored electronic copies of past communications. By matching the perspec-tive of the law to the perspective of the police, both courts made sure that theFourth Amendment’s set of rules applied to the officers’ interactions with theInternet just as it would with the officer’s interactions with the physical world.

2. Mediating Between the Two Approaches and the Need for Judgment

I have now articulated two ways of resolving the problem of perspective:First by examining the doctrine to see if it reflects an implicit or explicit choiceof perspective, and if that fails to produce an answer, then by following theperspective of the party that the applicable law seeks to regulate. Although thesetwo guideposts can help resolve the problem of perspective, they are notself-executing. Difficult questions of judgment remain. This subsection exam-ines some of the difficult questions that can arise when we apply the frameworkto different cases.

One difficult problem that arises in applying the framework is this: How doyou know when the first approach resolves which perspective to apply, obviat-ing the need for the second? Consider the Fourth Amendment. At first blush, itseems plausible to conclude that the modern Fourth Amendment simply reflects

228. The officers also alleged statutory claims but were unsuccessful. Id. at 1236.229. Id. at 1234 (emphasis added).


an internal approach, or alternatively tasks us with following the perspective ofthe person whose privacy rights are at stake. After all, the touchstone of themodern Fourth Amendment asks whether the government conduct violates a“reasonable expectation of privacy.”230 This certainly sounds like an internalinquiry. The doctrine focuses on the user’s expectation, and in particularwhether the user’s expectation is reasonable. Further, it is generally understoodthat the Katz case that introduced the “reasonable expectation of privacy” testalso rejected the Olmstead majority view (which featured Taft’s external ap-proach) in favor of something closer to Brandeis’s dissent (which adopted aninternal approach).231 If this is true, then the first approach would tell us tofollow the user’s perspective, and we should not get to the second approach,which pointed us to the police officer’s perspective.

I think the resolution of this tension lies in a close reading of the casesinterpreting the Fourth Amendment. A majority of the Supreme Court hasgenerally rejected the notion that an expectation of privacy is “reasonable”merely because a reasonable person would expect privacy in the circum-stances.232 Instead, the Court has required that the expectation be “legitimate,”which often leads the court to an analysis heavily influenced by whether thegovernment has violated the defendant’s property rights in the course of itsinvestigation.233 This approach has often focused the courts (rightly or wrongly)on external rather than internal questions when it has applied the FourthAmendment to new technologies. Ironically, despite the common wisdom thatthe Supreme Court has rejected Taft and embraced Brandeis, the Court’sdecisions more closely resemble Taft’s external approach than Brandeis’s inter-nal method.

The clearest example of this appears in Smith v. Maryland.234 In some ways,Smith offers a repeat of the arguments in Olmstead, except this time post-Katz.Smith considered the Fourth Amendment implications of installing a pen regis-ter, which is a device that records the telephone numbers dialed on a particularphone.235 The police asked the phone company to install a pen register on thetelephone line of Michael Lee Smith to confirm that Smith was a stalker thepolice were trying to catch.236 When Smith later called his victim’s home, thepen register attached to his phone line at the telephone company recorded the

230. Katz v. United States, 389 U.S. 347, 361 (1967) (Harlan, J., concurring).231. See, e.g., LESSIG, supra note 20, at 117 (“The Court in Katz followed Brandeis rather than

Taft.”).232. I developed this at length in Kerr, supra note 57, at 507–12.233. See id. See generally Thomas K. Clancy, What Does the Fourth Amendment Protect: Property,

Privacy, or Security?, 33 WAKE FOREST L. REV. 307 (1998); Orin S. Kerr, The Fourth Amendment andNew Technologies: Constitutional Myths and the Case for Restraint (unpublished manuscript, on filewith author).

234. 442 U.S. 735 (1979).235. See United States v. Guglielmo, 245 F. Supp. 534, 535 (N.D. Ill. 1965).236. Smith, 442 U.S. at 737.


fact that he had dialed her number.237 The police then used this information tosupport a warrant to search Smith’s home, which resulted in evidence that led toSmith’s conviction.238 Before the Supreme Court, Smith contended that the useof the pen register violated the Fourth Amendment.239 In effect, Smith offeredan internal argument: He reasoned that from his perspective as a telephone user,his dialing of the numbers was a private act.

In an opinion by Justice Blackmun, a majority of the Court applied anexternal perspective of the telephone network to reject Smith’s claims.240 Themajority did not look to whether dialing phone numbers seems private to atelephone user. Instead, the court focused on how the telephone network workedand how the network processed the dialing of numbers.241 Echoing Taft’sreasoning in Olmstead, Justice Blackmun treated Smith’s act of dialing thenumbers as the functional equivalent of telling an operator the number to bedialed:

When he used his phone, [Smith] voluntarily conveyed numerical informationto the telephone company and “exposed” that information to its equipment inthe ordinary course of business. In so doing, [Smith] assumed the risk that thecompany would reveal to police the numbers he dialed. The switching equip-ment that processed those numbers is merely the modern counterpart of theoperator who, in an earlier day, personally completed calls for the subscriber.[Smith] concedes that if he had placed his calls through an operator, he couldclaim no legitimate expectation of privacy. We are not inclined to hold that adifferent constitutional result is required because the telephone company hasdecided to automate.242

Notably, the dissenting justices found this external approach wanting. Theyoffered an internal perspective, from which the majority’s external analysis “nomore than describe[d] the basic nature of telephone calls”243 and ignored theprivacy interests of telephone users when they were placing calls.244

I do not think Smith v. Maryland resolves the question of whether the FourthAmendment is internal or external. However, it does reveal some of thecomplexities of choosing between real and virtual perspectives. The phrase“reasonable expectation of privacy” does not itself determine which perspectiveto apply. Instead, if our goal is fidelity to existing law, we have to conduct acareful inquiry into how the courts have applied the test in prior cases. Whenthe results of that test prove inconclusive, as I think they do here, following the

237. Id.238. Id.239. Id. at 742.240. Id. at 741–46.241. See id. at 742–43.242. Id. at 744–45.243. Id. at 746 (Stewart, J., dissenting).244. Id. at 747–48 (Stewart, J., dissenting).


perspective of the regulated party (as the Court did in Smith245) can provide ananswer. We can probably look at this problem through the lens of the firstapproach as well, by concluding that the doctrine points us to follow theperspective of the regulated party. In the case of the Fourth Amendment,doctrine from precedents such as Smith, Charbonneau, and Bohach suggest thatwe should follow the perspective of the police officers conducting the investiga-tion.

We can see another twist on this problem by recalling the second hypotheticalfrom Part II. This problem considers whether a search warrant authorizing asearch of “123 Pine Street, Suite 200” allows officers to search for and retrievefiles that are virtually present on the network but actually located on a remoteserver far away.246 From an internal perspective, the officers can search andretrieve the files because they are virtually present; from an external perspec-tive, they cannot conduct the search because the files are actually located in adifferent location. Which perspective wins?

In this case, I think the doctrine does provide a clear enough answer so thatresort to the perspective of the police is not necessary. Rule 41 of the FederalRules of Criminal Procedure makes clear that search warrants can be authorizedto conduct searches only in specific places in the physical world.247 Althoughnot a constitutional command,248 the Rule limits search warrants to searches ofproperty within the federal judicial district in which the warrant is obtained,with limited exceptions:

a search warrant authorized by this rule may be issued (1) by a federalmagistrate judge, or a state court of record within the federal district, for asearch of property or for a person within the district and (2) by a federalmagistrate judge for a search of property or for a person either within oroutside the district if the property or person is within the district when thewarrant is sought but might move outside the district before the warrant isexecuted and (3) in an investigation of domestic terrorism or internationalterrorism, . . . by a Federal magistrate judge in any district in which activitiesrelated to the terrorism may have occurred, for a search of property or for aperson within or outside the district.249

This language shows careful attention to the location of the property that can

245. In Smith, the pen register information was simply data that the phone company could collect tohelp the officers gather probable cause to obtain a warrant to search Smith’s office. The police were not“online” in “phonespace.” Accordingly, their perspective was external, as was that of the majority. Seeid. at 741–46.

246. See subsection II.A.1.247. See FED. R. CRIM. P. 41.248. It may, however, have constitutional dimensions. See Weinberg v. United States, 126 F.2d 1004,

1006–07 (2d. Cir. 1942) (holding that a search warrant issued by a District Court judge in Michigan toseize property in New York violates both “[t]he Fourth Amendment and the legislation governingsearch warrants”).

249. FED. R. CRIM. P. 41(a) (emphasis added).


be searched by a particular warrant. It refers specifically to when a warrant canauthorize a search “outside the district.”250 Given that these districts are federaljudicial districts, not more ethereal concepts that could include a “district ofcyberspace,”251 the Rule suggests that the permissibility of a remote searchshould be resolved through an external perspective.

A district court recently applied such an external approach to a remote searchin United States v. Gorshkov.252 In Gorshkov, FBI agents in Seattle conducted aremote search of an Internet account in Russia without a warrant to secureevidence of computer hacking conducted by a Russian suspect who had justbeen arrested and taken into custody in Seattle.253 The Russian hacker chal-lenged the warrantless search on the ground that it violated his Fourth Amend-ment rights. In rejecting the hacker’s argument, the court applied an externalperspective to the search; it viewed the search as a search occurring in Russia,not one occurring in Seattle or in cyberspace.254 Although Gorshkov involves awarrantless search, its application of an external perspective to a remote searchover the Internet suggests that the territorial implications of Internet searchesshould and will be resolved primarily through an external perspective.255

A final wrinkle in applying the framework arises when doctrine does notanswer which perspective to apply, and we turn for guidance to the perspectiveof the party to be regulated. The difficulty here is that perspectives may change;a regulated party may switch perspectives, changing from an internal to anexternal perspective or vice versa in the middle of the conduct that the lawseeks to regulate. For example, a law enforcement agent executing a searchwarrant to seize a computer file from a network may approach the networkexternally at first (entering the building where the server is located), but thenswitch to an internal perspective (by logging on to the network to access thefile). This poses a difficult problem: If we follow the perspective of the

250. Id.251. See Henry H. Perritt Jr., Jurisdiction in Cyberspace, 41 VILL. L. REV. 1, 100–01 (1996)

(speculating that someday there may be a “U.S. District Court for the District of Cyberspace”).252. No. CR00-55OC, 2001 WL 1024026 (W.D. Wash. May 23, 2001).253. See id. at *1.254. See id. at *3 (“The Fourth Amendment does not apply to the agents’ extraterritorial access to

computers in Russia and their copying of data contained thereon. First, the Russian computers are notprotected by the Fourth Amendment because they are property of a non-resident and located outside theterritory of the United States.”) (citing United States v. Verdugo-Urquidez, 494 U.S. 259 (1990)).

255. An interesting line of cases that may challenge this conclusion are those cases holding that thepolice may answer telephone calls inside a house when executing a search warrant inside that house, atleast so long as there is a good chance that the telephone call will yield additional evidence of thecrime. See United States v. Gallo, 659 F.2d 110, 113–14 (9th Cir. 1981); United States v. Fuller, 441F.2d 755, 760 (4th Cir. 1971). If the police may answer the telephone in the place to be searched, theyarguably can accept incoming data sent over the Internet to the place to be searched as well. Thedifficulty with this argument is that in the case of an Internet search, the search occurs when the officersends computer commands to the remote server ordering it to look through the server and send filesback. It is the search of the server, not the receiving of the data, that raises Fourth Amendmentdifficulties.


regulated party, do we switch perspectives as the party switches from offline toonline and perhaps even back?

I think the best approach is to apply the primary perspective that the regulatedparty embraces as it approaches the Internet. Parties should not be allowed tocircumvent a regulatory scheme by shifting perspectives to trigger a lessburdensome test. It would gut Fourth Amendment protections to allow police toswitch perspectives in a conscious effort to trigger a lower burden. Courtsshould not allow this.256 What matters in the case of the Fourth Amendment iswhether the officer interacts with the Internet as a virtual world (as in the chatroom of Charbonneau) or as a physical network (as in the server storingmessages in Bohach). The primary perspective should govern.

CONCLUSION

The problem of perspective sheds light on the law of the Internet by revealinghow several problems of cyberlaw are really conflicts between internal andexternal renderings of the facts. This insight does not answer which perspectiveto apply. However, I think it does raise doubts about some of the granderaspirations of cyberlaw. Whether the problem of perspective will remain asimportant as it is today depends on two competing forces. On the one hand, theadvance of technology will make virtual realities more lifelike and convincing.This should make the internal perspective more appealing, at least to some; thejump from physical reality to virtual reality will seem less forced and moreintuitive. At the same time, I expect that the external perspective will drawadherents from the increasing numbers who understand how the Internet works.As more people understand the technical details of the Internet, those detailsmay become part of the public’s understanding of the technology, fueling theexternal perspective. Perhaps one of these influences will overpower the other,establishing a more internal or external approach over time. Or perhaps anequilibrium will be reached, and both perspectives will survive and continue toshape the law of the Internet in the future.

256. Although recent Fourth Amendment jurisprudence shies away from inquiring into the subjec-tive intent of the officer conducting the search, see, for example, Whren v. United States, 517 U.S. 806,813 (1996), significant exceptions still exist. See Indianapolis v. Edmond, 531 U.S. 32, 44 (2000)(allowing inquiry into the programmatic purpose of a roadblock scheme to help determine its constitu-tionality); Florida v. Wells, 495 U.S. 1, 4 (1990) (stating that “an inventory search must not be a rusefor a general rummaging in order to discover incriminating evidence”). Accordingly, the law would notclearly foreclose an inquiry into whether the officer switched perspectives to be regulated by a lessdemanding standard.


ssrn id310020

Technology

civil liberties union

federal magistrate judge

virtual storage locker

gulf shipbuilding company

record companies copyrights

fourth amendment restrictions

organized crime activity

infra section iv