Andrew Lindemann Malone's Internet Playpen
Movie Reviews

Is Surfing the Internet Like Buying a Magazine or Tuning in a Channel?

Reasoning From Precedent and Metaphor in Reno v. ACLU (1997)


I. Introduction

Legal reasoning has always depended on metaphor to make abstract concepts real, even when the reasoning itself does not explicitly acknowledge the metaphors it employs. From the famous "marketplace of ideas" to the idea of "choking off the flow" interstate commerce, lawyers, judges and scholars have depended on the power of metaphor to make abstract ideas both more real and more easily manipulated and debated.

But one area of legal reasoning whose dependence on metaphor has not been much examined is the process of reasoning from precedent. While reasoning from precedent does not always entail the use of metaphorical reasoning, there are certain circumstances in which it is impossible to characterize competing arguments from precedent as anything but competing metaphorical characterizations of the case presently under consideration.

In this paper, I will give a brief characterization of reasoning from precedent using blending theory, and explain under what circumstances reasoning from precedent about a case might be considered metaphorical. I will then examine one such case: the 1997 Supreme Court decision Reno v. ACLU. This was one of the first Supreme Court cases involving the Internet, meaning that there were no close precedents on which the Court could rely. The central arguments which are addressed in the decision are therefore arguments about the best way to metaphorically characterize the Internet in terms of cases which could establish precedent. I will also examine the explanatory power of the metaphors, and speculate with regard to motivation for judging a metaphor as valid or invalid in legal reasoning.


II. Reasoning from precedent, blending theory, and analogy vs. metaphor

In the Anglo-American legal system, precedent is the gold standard for legal reasoning. We use precedents so that people will have reasonable expectations about what they can and cannot do, so that people in similar situations are treated similarly by the law, and so that the legal system itself is self-consistent and therefore achieves legitimacy in the eyes of the people. Therefore, it is critically important to come to an understanding of how exactly we reason from precedent.

At first glance, reasoning from precedent may appear merely to involve finding already-decided cases which bear some relation to the case presently under consideration, and repeating the decision from those earlier cases in the present case. Yet this does not characterize fully the factors that enter into argument from precedent, and which those who argue from precedent must take into account. The argument from precedent must also address the language of the decision in the cases being used as precedents, the language of the law under which the previous case was decided, and any salient characteristics of these previous cases which may or may not have a bearing on the case presently under consideration.

Blending theory has proven useful for providing a framework which can integrate inputs from many spheres of knowledge, and it can give a good account of the process of reasoning from precedent. I will present a simplified account of a generic case here, since in reality arguing from precedent involves many more inputs than we need account for here. But the basic principles are the same.

All statutory legal disputes start with some kind of law.1 A law is supposed to provide guidelines which will allow a judge to make a decision about the legality or illegality of an action. But no law attempts to take into account every characteristic of the actions it declares illegal; rather, it described a type of action, which is necessarily generic. It will not take into account, for example, the hair color of the person (or group, corporate entity, etc.; for purposes of this discussion we will assume the actor is a single person) who performs the action. More seriously, it may not specify the gender of the person who performs the action, or the motive, or even the means. The law will only specify what its makers think are the salient characteristics of the action they are trying to outlaw.

In blending theory, we call the different realms of knowledge which provide material for our reasoning "input spaces." Therefore, in an account of reasoning from precedent, an input space for the language of the law itself will have to be created, as in Figure 1.



Now let us assume that the law has just been created, and a court is now trying the first person accused of breaking that law. Let us further assume that this person is guilty of breaking the law. The input space which characterizes his action will contain all of the characteristics of the action, and also, as discussed above, some characteristics which are not discussed in the law itself. This person's input space is shown in Figure 2.



Blending theory states that, given these two input spaces, we recognize their shared attributes and conceive of a generic space which contains the elements both input spaces share, i.e., characteristics #1, #2 and #3.

The interesting space here is the blend space, which we will call the judgment space and use to represent the mental space in which the actual decision of guilt or innocence is made. The judgment space recognizes the correlations between the shared characteristics of the action described by the law and the action performed by the lawbreaker. But in making the judgment and pronouncing this lawbreaker guilty, the judge will also provide reasons for recognizing the correlations and making a guilty judgment. In blending theory, this reasoning would be called an example of completion, the additional information we supply to create a "full" blend space. Similarly, any inferences drawn from this reasoning about what types of people might or might not be judged guilty in the future under this law would be called extensions in blending theory. All these things combine to make up the judgment space, as shown in Figure 3. (The generic space has been omitted for simplicity.)



This case establishes a precedent. There is a law, a case which has been interpreted by the law, and an explicit statement of the reasoning behind that interpretation. The stage has been set for reasoning by precedent the next time someone is accused of breaking this law.

The key thing to recognize here, again, is that reasoning from precedent is affected not only by the previous cases decided under the law, but also on the reasons given for the decision, the language of the law itself, and even the inferences the judge may have drawn in deciding the first case. The reasoning by precedent to make a decision on a second person's action is shown in Figure 4.

Obviously, not all reasoning which can be characterized by blending theory is metaphorical reasoning. This example would probably be called analogical reasoning by most people. But blending theory allows us to understand that the cases, the law and the reasoning which go into making a decision are all from separate spheres of thought.




This example concerns correspondences which are motivated by obvious resemblances: outward characteristics which we have assumed are nearly identical. But there exist cases for which reasoning by precedent necessitates inventing some creative relationships between characteristics of the previous cases, the law itself, and characteristics of the case presently under consideration. When the input spaces are different enough from each other, in normal reasoning and in the law, we characterize the relationship between them as metaphorical. There is no systematic way to draw the line between analogy and metaphor; such distinctions are mostly based on feelings people have about the way reality is structured. Still, the reasoning involved in judging some cases seems clearly to belong to metaphor, and Reno v. ACLU is one of these cases.


III. Background of Reno v. ACLU

The disputed law, known as the Communications Decency Act (CDA), was part of a larger bill called the Telecommunications Act of 1996 which attempted to reform telecommunications for the Internet age. As Justice John Stevens acidly notes in his majority opinion, "The Act includes seven titles, six of which are the product of extensive committee hearings and the subject of discussion in Reports prepared by Committees of the Senate and the House of Representatives. By contrast, Title V [the CDA] contains provisions that were either added in executive committee after the hearings were concluded or as amendments offered during floor debate on the legislation."

Two provisions of the CDA were immediately challenged on First Amendment grounds by a diverse collection of plaintiffs, led by the American Civil Liberties Union and including Internet providers, anti-censorship groups, and groups who thought their speech might be silenced by the new law. The first of these provisions, referred to either as Section 223(a) or as the "indecent transmission" provision, criminalized the "knowing" transmission of "obscene or indecent" content to persons under 18 years of age, "regardless of whether the maker of such communications placed the call or initiated the communication." The second provision criminalized the use of "any interactive computer service to display in a manner available to a person under 18 years of age" any content which "depicts or describes, in terms patently offensive as measured by contemporary community standards, sexual or excretory activities or organs, regardless of whether the user of such service placed the call or initiated the communication."

The Court held that "the CDA's 'indecent transmission' and 'patently offensive display' prohibitions abridge 'the freedom of speech' protected by the First Amendment." Seven of the court's members, for whom Justice Stevens wrote an opinion, elected to strike down the law altogether; Justice Sandra Day O'Connor wrote another opinion, with which Chief Justice William H. Rehnquist concurred, using slightly different reasoning to preserve part of the law.

As might be guessed, the CDA had some problems from a legal standpoint which were not discussed metaphorically. The lack of congressional scrutiny for the CDA was noted in Stevens' opinion, as quoted above. There was also some skepticism about establishing a regulatory role for the government in a sphere where it had historically allowed free play, and the law's definitions of "indecent" and "obscene" were judged to be inadequate in a legal context. However, both Justices Stevens and O'Connor spend much of their opinions addressing the precedential arguments put forth by the government using metaphorical reasoning, and it is with this reasoning that we concern ourselves now.


IV. Metaphorical argument in Reno v. ACLU

We have seen that reasoning from precedent sometimes uses metaphorical reasoning, especially if the case one is attempting to employ as a precedent and the case presently under consideration are unusually dissimilar. Given that, it is interesting to note that neither Justice Stevens nor Justice O'Connor explicitly or implicitly acknowledges that the arguments about precedent which they address are arguments about the fitness of various metaphors. Justice Stevens makes a point of noting that the Internet cannot be characterized in terms of other media, writing in his opening statement of the "undisputed facts" of the case that the Internet is "a unique and wholly new medium of worldwide human communication," and noting a few paragraphs later that the Internet is "a unique medium–known to its users as 'cyberspace'–located in no particular geographic location but available to anyone, anywhere in the world, with access to the Internet." These comments would seem to imply that any reasoning from precedent here would have to be metaphorical, since the uniqueness of the Internet would preclude what we think of as analogical correspondences between precedents and the present case.2 However, Justices Stevens and O'Connor did not explicitly acknowledge that they were using metaphoric reasoning.

Nevertheless, it seems clear upon analysis that the Court used metaphoric reasoning to dismiss the three precedents the government submitted as validating the CDA, and eventually to select the case which the Court found was most relevant to the matter at hand. In arguing, the government relied on precedents from three cases in which the Supreme Court had upheld limits on access to indecent speech.3 In Ginsburg v. New York (1968), the Court held that a law barring minors under 17 from purchasing material containing indecent speech was constitutional. In FCC v. Pacifica Foundation, the Court held that the Federal Communications Commission had a right to sanction broadcasters who broadcast indecent speech. In Renton v. Playtime Theaters (1986) the court upheld a zoning law which outlawed locating adult movie theaters in residential neighborhoods. Justice Stevens states flatly that a "close look at these cases…raises–rather than relieves–doubts concerning the constitutionality of the CDA."

Stevens then proceeds to explore the differences between each of the precedents cited by the government and the case presently under consideration. This reasoning can be characterized by conceptual metaphor theory; Stevens sets up a model of the situation "source case," or precedent, and then attempts to map it onto the "target case," Reno v. ACLU. For example, a model of the mapping between Ginsburg and the present case, and the entailments of that mapping, looks like this:





Webmaster/Internet provider


Web surfers/Internet users

places of business



downloading, viewing, etc.


proof of identity

checking ID

verifying proof of identity

physically present pornography

pornographic Internet content

totality of available content

totality of Internet content


This mapping can be seen as the test the government's cited precedents must pass–if any important part of the mapping is invalid, then the source case is not valid as a precedent. In addition, if the mapping leaves out any important information, the source case is not valid. It is especially helpful to think of reasoning from precedent in the context of blending theory here, as it helps us recognize that we bring knowledge of two fully formed environments to the metaphoric mappings. It could also be the case that a case might not be valid as a precedent even if the mapping were valid, because using it as a precedent might ignore the reasoning involved in originally deciding the case. The latter situation does not arise in this case.

This mapping is of course speculative, because Justices Stevens does not specify the underpinnings of his reasoning, but it explains his objections to the use of this precedent. For example, as Stevens notes, "the New York statute applied only to commercial transactions…whereas the CDA contains no such limitation." Thus he denies the metaphorical correspondence between buying and viewing. Earlier, in the section which relates "undisputed facts," he quotes the District Court as having determined that there "is no effective way to determine the identity or the age of a user who is accessing material through e mail, mail exploders [listservs], newsgroups or chat rooms," meaning that Justice Stevens would dispute the assertion, implied by the use of the Ginsburg case as precedent, that the age of an Internet user could be easily determined to ensure compliance with the law. The correspondence between checking an ID and verifying an identity online is therefore rejected. Stevens also addresses omissions in the mapping by noting that there is a situation accounted for in the Ginsburg decision, that where a parent purchases indecent material for his or her child, which is not accounted for in the CDA, which does not take into account parental consent. All these assertions can be seen as a declaration that the metaphoric mapping between the Ginsburg decision and the CDA is invalid.

The other two precedents cited by the government are dismissed in a similar manner. Here is a speculative mapping relating Pacifica and Reno v. ACLU:





Webmasters/Internet providers


Web surfers/Internet users




downloading, viewing, etc.

frequency (station)

Web address, chat address, etc.

tuning in a station

surfing to a website

moving up and down the dial

random surfing


This mapping has a number of flaws in Justice Stevens' eyes. Chief among them is the difference between the two mediums in terms of the process of accessing content. In another passage in the "undisputed facts" section, he quotes the District Court as having found that "the receipt of information on the Internet requires a series of affirmative steps more deliberate and directed than merely turning a dial." Radio and television had received the most limited First Amendment protection "in large part because warnings could not adequately protect the listener from unexpected program content." In contrast, "[a]lmost all sexually explicit images [on the Internet] are preceded by warnings as to the content." Tuning in a station, in Justice Stevens' view, is not even remotely equivalent to surfing to a website or going to a chat address.

Furthermore, the mapping above omits a lot of important information. Radio has a long history of governmental oversight; the Internet does not. The disputed FCC order in Pacifica was not punitive; the CDA imposes penalties on those found in violation of it. Finally, the content regulation in Pacifica involved what is called a "time, place and manner analysis," which means that indecent speech may be permitted at one time and not at another (which is, in fact, current FCC policy). The CDA bans everything indecent at every time in every manner. Still, even if the CDA had been modified somewhat to be more in line with the Pacifica ruling (although it is hard to see how this could be done), it still seems that Justice Stevens would not have accepted the precedent because of the above metaphorical breakdown between tuning in a station and surfing to a website, especially since this problem is mentioned in at least three separate places in the decision.

The third precedent is somewhat nonsensical, and it is hard to imagine it being seriously advocated as a precedent in front of the United States Supreme Court. It produces a mapping like this:






Web address


markers of certain types of sites

residential zones

zone free of indecent speech

secondary effects (crime, deteriorating property values)

slower growth of the Internet


The CDA, of course, does not actually have the elimination of secondary effects as its goal, but rather the "primary effects of 'indecent' and 'patently offensive' speech," as Justice Stevens writes. In addition, while in real life there are residential zones and other types of zones, under the CDA the entire Internet would be a zone free of indecent speech. Neither of these metaphorical correspondences is valid. In addition, the government's argument that the CDA is necessary to foster a "secondary effect" of growth of the Internet ignores the fact that the Internet had been growing spectacularly quickly without the interference of the CDA.

Still, the idea of viewing the Internet as a geographical entity is an intriguing one, and it is this idea that Justice O'Connor examines in her partially dissenting opinion. She borrows the idea of geographical zoning, but otherwise does not address the claim that Renton is a precedent for Reno v. ACLU, presumably because she ascribes no validity to it. In her judgment, the two features of the physical world which allow us to zone it and keep minors out of certain zones are geography and identity. She refers to a number of Supreme Court cases which she asserts have the same basic pattern, and extracts the basic pattern by induction. She suggests an adult dance show as an example, and it may be instructive to do a mapping between such an establishment and the Internet:




theater operator

Webmaster/Internet provider


Web surfer/Internet user

location of show

Web address, location of chat room, etc.

area around show

free access zone

show itself

restricted access zone

ticket taker

identification verifying software


verifiable identification


Justice O'Connor refers to a fact found in the District Court to claim that "cyberspace 'is moving…from a relatively unzoned place to a universe that is extraordinarily well zoned'", and explicitly states that "chat rooms and Web sites, for example, exist at fixed 'locations' on the Internet." She ends up rejecting the argument from these precedents (which, it should be noted, were not suggested as precedents by the government) because it is impossible to establish identity in cyberspace in the same way that one can in the physical world, and thus "a speaker cannot be reasonably assured that the speech he displays will reach only adults because it is impossible to confine speech to an 'adult zone.'" Justice O'Connor's eventual partial dissent revolves mainly around terminology and not around metaphor, but it is interesting to examine how she took a different metaphorical path than Justice Stevens did to arrive at the same conclusion: an examination of precedents suggests that the CDA (or, at any rate, part of it) is unconstitutional.

Since an argument from precedent is such a valuable thing to be able to make, one might expect one of the justices to attempt to evaluate the CDA under a precedent from some other case. Justice Stevens does in fact do this, likening the CDA to the ban on "dial-a-porn" services which the Supreme Court invalidated in Sable Communications of Cal., Inc. v. FCC (1989). "Dial-a-porn" is apparently a system whereby interested parties call a specific telephone number and hear prerecorded indecent speech. A mapping between the situations described in the two cases might look like this:




dial-a-porn operators

Webmasters/Internet providers


Web surfers/Internet users

answering machines



downloading, viewing, etc.

phone number

Web address, chat room location, etc.

calling the number

surfing to a Web site


This mapping resembles the mapping between radio/television stations and the Internet which was rejected above. The difference which makes this one valid and the other one invalid, in the view of the court as stated in Sable, is that "the dial it medium requires the listener to take affirmative steps to receive the communication…Placing a telephone call…is not the same as turning on a radio and being taken by surprise by an indecent message." Since the Internet also requires such affirmative steps, the correspondence between calling a telephone number and surfing to a Website is therefore valid, making this case usable as a precedent. Indeed, Stevens has prepared us to accept this metaphoric mapping  in the statement of undisputed facts, by quoting the District Court as having found that "[e]ach [Website] has its own address–rather like a telephone number."

At this point in the opinion, Justice Stevens has not quite established that the CDA is unconstitutional; rather, he has established that it is similar to another law restricting indecent speech which was unconstitutional, and dissimilar to other laws restricting indecent speech which were constitutional. The methods Justice Stevens uses to establish that the CDA is unconstitutional are not metaphoric at their base, but the metaphoric language he employs tells us something about his eventual conclusion.

In the findings of fact section of the opinion, the Internet had been characterized metaphorically for what seemed to be purely illustrative purposes. It had been called "both a vast library including millions of readily available and indexed publications and a sprawling mall offering goods and services," and, as noted, Web addresses had been compared to telephone numbers.

But the seeds were laid for metaphorical characterizations later in the argument in the section discussing undisputed facts. The Internet was also referred to as "a vast platform from which to address and hear a world wide audience of millions of readers, viewers, researchers and buyers." This characterization of the Internet, as a publishing medium, becomes the predominant characterization in Justice Stevens' opinion after the government's precedents have been dismissed. Justice Stevens compares the government's position "to arguing that a statute could ban leaflets on certain subjects as long as individuals are free to publish books." In the very last paragraph of the opinion, Justice Stevens calls the Internet a "new marketplace of ideas," granting it by extension the fullest protection under the First Amendment. Even if one knew only of the metaphors Justice Stevens uses for the Internet in the final pages of his opinion, one could easily infer that he would oppose any attempt to regulate its content.


V. Motivation and conclusion

Obviously, the motivation for metaphors used in reasoning from precedent is normally one of a generic-level resemblance; two differing situations are nevertheless perceived to have a similar basic structure, and therefore a decision which applies to one should apply to the other. Resemblance, however, is in the eye of the beholder, and in fact many uses of metaphorical reasoning in reasoning from precedent seem designed not so much to explore the true legal import of the case presently under consideration as to provide ammunition for a view which the interested party wishes to prevail. In other words, like the drunk and the lamppost, metaphorical reasoning is used to provide support and not enlightenment.

Is this the case in Reno v. ACLU? In my opinion, it is not. The three precedents suggested by the government do, in fact, seem to have metaphorical invalidities in them which make them useless for reasoning from precedent. The idea that Sable is a better precedent is more debatable. It is no doubt true that accessing indecent speech on the Internet requires more "affirmative steps" than turning on a radio, but I would assert that in many cases calling a "dial-a-porn" telephone number requires more such steps than accessing an Internet site does. There are, of course, other key distinctions between Sable and Pacifica which make Sable a better precedent for this case, most notably the fact that the law being considered in Sable provided a total ban while that in Pacifica did not.

It seems fair to say that to some extent, Sable is invoked simply because it is wise to refer to precedent in legal opinions. The CDA would not have stood in any case; Justice Stevens' majority opinion makes it clear that there are enough technical problems with the law that it could not have passed First Amendment scrutiny. But the fact remains that the process of making sure that the CDA could be exposed to the highest level of First Amendment scrutiny required the justices to reject the government's reasoning from precedent, and because of the unique nature of the Internet, an evaluation of these precedents required metaphor to be effective. A greater appreciation of the role of metaphor in reasoning from precedent will allow us to gain a fuller picture of how metaphor works in shaping the laws and court decisions by which we live.


1Common-law disputes, of course, do not  start with an explicit law. I have used the statutory case as an illustration because it is simpler to explain with blending theory, but blending theory can be made to work for common-law disputes as well.

2Interestingly, one of the justices for the District Court that entered the preliminary injunction against enforcement of the CDA, Justice Dalzell, actually argued that past Supreme Court cases required a "'medium specific' approach to the analysis of the regulation of mass communication," but this argument is ignored after it is quoted.

3Indecent speech is different from obscene speech, which has no constitutional protection; the distinction between the two types is, to say the least, not ideally clear.


All this tasty writing ©2002-11 by Andrew Lindemann Malone. All rights reserved.