The Hyperlinked Society: Questioning Connections in the Digital Age

DAVID WEINBERGER

The Morality of Links

Links are good. I believe that. And I'm not indifferent to the statement, the way I am to the vast majority of facts with which I agree, such as “There is a worm somewhere in our front lawn” and “Venus Williams plays tennis better than I do.” The Web, as an infrastructure built of links, brings me joy and even an occasional sip of hope. I just don't know exactly what it means to say that links are good.

1.

Atomic bombs are bad. Not every conceivable use of atomic bombs is bad. There is a reasonable case to be made that using one to kill one hundred thousand civilians in Hiroshima was a good thing to do. We don't have to agree about this. I personally do not think it was, but I'm not as certain as I once was. Even so, that nudging of my position about the bombing of Hiroshima¹ has not nudged my belief that atomic bombs are bad.

As everyone knows, technology is not good or bad in itself. The iron maiden, that instrument of medieval torture, would be mighty handy if you tipped the spikes with life-giving medicine and gently closed the door part way on a patient of some sort. Or perhaps they would make the perfect planters for an herb that can save lives. Or make up your own example. Despite the ability to dream up good uses for iron maidens, I'm still willing to say they're bad.

But a funnel is not bad. Why, I've used one myself, to pour oil into a car engine. And if you respond that given the fuel inefficiency of car engines, funnels are bad, then let me tell you about the time I used a funnel to filter out the cork remnants from a delightful but cheap wine, to be drunk in moderation. And if delightful but cheap wine is not good, then nothing is. But funnels are also used in a form of “waterboarding,” a contemporary form of torture sanctioned by my government. Even so, my belief that funnels are not bad remains unshaken.

Likewise, I do not believe rocks, sticks, or bricks are bad, although they have been used to commit unspeakable and at times unimaginable acts of evil. The difference between funnels, rocks, sticks, and bricks, on the one hand, and iron maidens and atomic bombs, on the other, is obvious. The difference does not lie in the intention of the creators, for those intentions don't adhere to the objects. Besides, what was the intention of the creator of rocks and sticks? Rather, the difference lies in what the social psychologists call affordances, that is, what these objects enable you to do. Funnels and bricks let us do many, many good things. Funnels let us transfer materials with little spillage. Bricks let us build structures that keep us warm and safe. Sticks and rocks have so many affordances that it's silly to even try to list them. But atomic bombs and iron maidens have fewer affordances. Their main ones result in pain and destruction. They're bad, even if we're just using an atomic bomb to prop open a door so an ambulance can get through. Their good uses are the exceptions.

If you see a problem in the preceding argument and it really bothers you, then you're a philosopher, whether you're willing to admit it or not. A philosopher reads the preceding and sees essentialism lurking beneath the surface. Essentialism, a doctrine that springs officially from Aristotle, picks out one definition—or meaning or use—among the many possible and gives it ontological priority. Essentialism says that there are many ways an atomic bomb can be used but that the real way is to blow things up. Essentialism says that a funnel can be used as a hat for the Tin Man but that its real use is to transfer pourable materials. In the previous paragraph, when I picked out some affordances over others, I was evaluating the moral character of objects by choosing one particular affordance over others. But since we no longer believe in essentialism—who is the privileged person or Person who gets to choose which use is the real one?—my assignment of moral properties to things is really just a shorthand for saying that we usually use funnels for good purposes and usually use (or intend to use) atomic bombs for bad purposes.

I am not an essentialist. But I am also not so blinded by philosophy that I've become an idiot.² There is some sense in which the use of funnels for transferring materials is more important than their use as hats. Things are like words. They have meaning, but that meaning isn't single-voiced and exclusive. Even so, we almost always understand words and things well. We rarely run into a word that genuinely stumps us, and we rarely encounter things that we stare at blankly, saying: “What the heck is that? Is it a car part or something I can eat?” It happens, but it's rare. Nor do we ever run into a word or thing that we insist can mean only one thing. The word elevator usually refers to one of those up-and-down conveyances, but if someone talks about an elevator as part of an airplane wing or as a type of mood-altering drug, we don't stare blankly as if the person just said “Umphlitz.” Likewise, when our child uses a funnel as a musical instrument, we don't snatch it from the child's hands and send him or her for a time-out for violating the funnel's essence.

Things and words do not have meaning apart from us. But in their involvement with us, they have meanings that are neither as baked in, universal, and unalterable as essentialism would have us believe nor as arbitrary and willfully changeable as we sometimes would like to believe. Bring together the affordances of things, our needs and desires, and our ways of thinking, and you come up with an inevitable sense that although funnels aren't really hats or musical instruments, they can be used that way. We need a middle ground that lets us prefer certain meanings and uses but that acknowledges other meanings and uses—a middle ground that also lets us talk about how what we bring to the party can reveal what's true and real about the things we encounter.

Eleanor Rosch's work provides this middle ground. In the 1970s, she showed empirically that we make sense of the world around us by clustering meanings rather messily around prototypical examples. A robin or a sparrow is a prototypical bird, whereas a penguin or an ostrich is not. Penguins and ostriches are birds, of course, but they're not great examples of birds. They are part of a loose-edged cluster of birds that's formed not around a definition but around examples as clear as robins and sparrows. Likewise, a funnel is a cluster of meanings and users, some of which are prototypical—pouring motor oil, transferring cooking ingredients—while others are out at the edge of the cloud of meaning. Over time, a funnel might become prototypically something we wear or blow through, just as salt has gone from a precious substance to the most common of edible commodities. And perhaps someday atomic bombs will be used primarily as a way to dig canals or power spaceships. At that point, I'll say atomic bombs are good, and we'll all chuckle about how we were once so primitive that we actually thought they were bad. But now isn't later. Now atomic bombs are bad.

How do we know that? As usual, that question really comes down to how we can explain ourselves if someone disagrees. After we've spent hours arguing and have started to pick up the empties, the fundamental question is this: is the world better off for having atomic bombs; or iron maidens; or funnels; or rocks, sticks, and bricks?

Or links?

2.

Links are more like sticks than like funnels. There are so many ways to use them that there are few prototypical uses. Perhaps the most prototypical is the Web page that links to another because it's on a related topic. But that's not much more prototypical than saying a stick is part of an old branch. In fact, it's worse than that. There are so many different reasons one page refers to another: to dig further into the same topic, to explore the topic more broadly, to explore a topic that's related but not the same, to see an example of a site that doesn't understand the topic at all, to get further evidence that what the page says is right, to propitiate an acquaintance, to get paid for running an ad someone clicked on. There are as many ways to link as there are to use a stick.

This occurred because Sir Tim Berners-Lee made sure there was only one way to link two pages. The HTML code that creates the link that shows up on the page as blue and underlined (typically) has no standard way of saying what the relationship is. A link to, say, www.martinlutherking.org is encoded in HTML as <a href=“http://www.martinlutherking.org”>MLK</a> and would show up on the Web page as a clickable “MLK” link. Nowhere in that code is there a place for the linker to note that www.martinlutherking.org is a hate site created by a racist organization called Stormfront.³ Berners-Lee's aim was to make linking as simple as possible. His success is evident in the hundreds of billions of links already created. If a Web page's author wants to explain the nature of the relationship to a page recommended, he or she can put an explanation into the text of the page: for example, “Here is a godawful, frightening, hateful page masquerading as a straightforward biography.” Whatever type of relationship an author can put into words can be expressed on the page. And when words don't work, the page can contain pictures and music. So the HTML code that expresses the link says nothing about the nature of the link, but the page that displays the link can say volumes.

Tim Berners-Lee has been working for years on a way of enriching the HTML code with more of the meaning of the link. That's called the Semantic Web. But we don't have to wait for it before we can talk about whether links are good or not, for links as they exist for humans are what's written on the page, not the HTML code intended to be seen and used by computers. The vast majority of those human links have a meaning—a semantics—that's at least somewhat explicit and obvious given their context; that is, the meaning is explicit and obvious to a human reader, even if a computer gets nothing more from them than a “Go fetch!” imperative.

We could perform an analysis of those links to get a sense of what sort of connections they're drawing. If the vast majority of them are embedded in text that consists of variations on “I hate that other site!” then my belief that links are good might be shaken, especially if the pages hosting those links were the most popular ones. But that's an unlikely outcome.

Besides—and more important—the goodness of links comes not from the quality of the pages they point to or the semantic contexts in which they're embedded. The goodness of links operates a level below that. So even if all the links on the Web were negative and hateful, I think I'd say not that links are bad but that there's something very nasty about us human beings. How else could we explain how we took something as useful as a stick and only figured out how to poke people in the eyes with it?

3.

How long have we been arguing over issues of abortion rights, gay marriage, capital punishment? How much longer do we have to continue before we'll just give up on the hope of resolving them?⁴ The length and ferocity of these arguments are strong empirical evidence that hotly fought moral debates cannot be settled.

This might lead us to despair, except we have equally strong empirical evidence that most moral issues do not need debate. Is it wrong to lie? Yes, unless you have a good reason. Is it wrong to punch someone in the face? Again, yes, except in special circumstances. Someone who lies for no reason and punches people in the face randomly isn't worth arguing with. He or she is only worth avoiding.

Our moral behavior and our ability to engage in moral argument are grounded in the same facts. You can't be moral if you don't recognize that there are other people with interests. If they're mere cartoons to you, even if you happen to be honest when making change and avoid slapping innocent babies, you're not yet moral. You're just a rule follower. To be a moral person—as you and I and just about everyone we know are—you have to be aware not only that there are others but that they care about what happens to them.

This assumption is built into the Golden Rule enunciated by Hillel the Elder and Rabbi Akiva⁵ and by a certain better-known rabbi who came after them. To imagine yourself in someone else's shoes is not just to imagine seeing the world from that person's point of view but, more important, to care about the world the way that person does. If all you can see is that other person's intellectual framework, then what you've understood is nothing. What counts is seeing how the world matters to the other person. That person would feel the pain your contemplated action might cause; he or she cares about the effect of your action on his or her career, children, guppy. This caring about what happens is at the heart of morality.⁶ Without it, morality becomes a mere set of rules. With it, the rules become rules of thumb we only consult when we have trouble sorting out the jumbled ways our actions matter.

Just a little unpacking of the Golden Rule reveals the obvious premises of moral life. If we want to see whether links are good per se and not just if they link to good stuff, we need to take a moment—here, a paragraph—to state the obvious.

The Golden Rule tells us that the possibility of morality itself depends on three fundamental facts: we share a world, that world matters to others, and the fact that it matters to others matters to us. If we remove any one of these three facts, the world isn't moral in any way we can recognize. Remove the first and what we do has no effect on others. Remove the second and it doesn't matter to others what we do to them. Remove the third and it doesn't matter to us what we do to others. (These are the views of solipsists, sociopaths, and really twisted sociopaths, respectively.) Put these three facts together and we live in a world in which our behavior is constrained because what we do affects others who also care about what happens to them.

Some moral principles can be derived from this infrastructure: it's good to consider the interests of others; it's good to try to understand others and what matters to them; it's good to let that understanding move us. These principles don't help us settle many disputes, because there are times when we need to frustrate the desires of others, and these principles don't tell us where the lines should be drawn. We should go back to the lesson that our world experience has taught us: we are never going to come to complete agreement, even as we slowly—oh so slowly—make progress as a species in, say, turning the tide against slavery and the subjugation of women.

But here I'm not trying to come up with universal laws of morality that can be trotted out to settle all the tough cases facing us. That would be useful if I were trying to decide whether most links are good or which types of links are good. But my purposes here are one level down. I want to see if there is anything about the structure and nature of links themselves that lets us say reasonably that links are good. Or, to be more exact, is there anything about their structure and nature that explains why at least some of us (I, for example) have a strong sense that links are good.

There is. If morality is based on our caring that we share a world with others to whom that world matters, then in acting morally, we turn toward that world with others. They point out to us that the world is this way or that and matters in that way or this. Making the comment “Hey, I'm trying to hear the movie!” reminds the person behind you that you're sharing a world that matters to you. Of course, it's unlikely to be very effective, since a person talking on a cell phone in a theater probably already knows you're trying to watch the movie but doesn't believe or doesn't care that his or her conversation is disturbing what should be your reasonable expectations for noise in a theater. Nevertheless, your instinctive comment leans on the right lever, trying to get your antagonist to see how the world looks and matters to you. In different circumstances, you might make more progress by engaging the person in conversation: “I'm just curious about why you think it's OK to take a call while in a crowded theater. Do you not know that we can hear you and it distracts us from the movie?” Maybe you won't get anywhere, but you're likely to get further than with your simple expostulation.

We sometimes make progress in morality by feeling the feelings of others, but we make more significant progress by understanding how the world appears such that it evokes those feelings. Sympathetic understanding is more powerful than mere empathy because it gets at more of the truth. Parents' grief for their child, for example, includes not just the universal grief parents feel but is embedded in an understanding of how the world occasioned the grief. Was it a wanton act of cruelty, part of a divine plan, a mere accident? These simple categories do not suffice. They merely sum up an event saturated with particularities. Our moral sense can go as deep as the world itself in understanding how things matter to those affected. Their grief is conditioned by and conditions all in their world. Everything matters differently.⁷ In this sense, then, morality is an infrastructure of connection in which we allow ourselves to care about how the world matters to others. That is formally the same as a description of the linked structure of the Web.

After all, what do we do on the Web? We link. No links, no Web. In linking, we send people to another site (assuming we aren't the sort of narcissists who link only to themselves) where they can see a bit of the world as it appears to another. We send our visitors to other sites because we think those other sites will matter to them. Our site probably explains why we think it will matter to them and how it matters to us, even if that explanation is “Here's a trashy site I hate.” Pointing people to a shared world, letting how it matters to others matter to us—that's the essence of morality and of linking.

4.

Morality and the Web have the same basic architecture? Holy Toledo! That means the Web is the same as morality. Surely the Web can't be that important. I must have slipped off the rails and crashed into the thickets of overstatement and Web utopianism.

I don't think I have. (But, then, of course I would think that.) It would be an overstatement if I were claiming that only the Web has this moral architecture. But despite the fact that Morality 101 is taught as a discrete course in college, the moral realm is not an isolated segment of human experience. If it were, it would indeed be a coincidence straining credulity if links happened to mirror its structure. But if morality is in fact the basis of human experience—or, to switch metaphors, if it permeates experience—then it's not too surprising that what we build for ourselves reflects that experience. Some of what we build reflects it more than others, but everything we build reflects it somewhat. It has to. We build things on purpose, and our purposes are always formed with awareness that we share a world. Outside of the odd cases where we build something purely for our own use and without a care for how it will affect others, we create in public and almost always for a public. If you manufacture funnels, you do so in order to help others achieve their purposes, thus implicitly acknowledging that we share a world with others who have interests.

But the moral structure of funnels is not in a funnel itself the way the Web's moral structure is in the Web. There is something special, but not unique, about the Web's moral architecture. The tools by which we communicate tend to reflect the moral architecture more explicitly than do funnels, sticks, and atomic bombs, because communication itself has the structure of morality: by communicating with each other, we turn toward the world that we share and that matters to both of us. I try to show you how the world matters to me, I attempt to understand how it matters to you, and we try to share more of the world. In communicating, I'm acknowledging not only that the world shows itself to both of us but that it matters to both of us. So as a communication medium, the Web is already structured morally.

The Web brings three new characteristics. They are not radically new—for what is?—but they are new enough to be worth noting. First, the Web is global in scope—and increasing its actual reach at a remarkable pace. Second, the Web turns the steep hill of broadcasting into a huge plain bordered by a cliff—once you're on, you're pretty much equal with everyone else, although if you're not on, you're pretty much off entirely. Third, the Web brings persistence not just to our communications but to the relationships our communications note; that is, the Web brings persistent links. In this, it is profoundly unlike other publishing media. Even if the cost of printing paper books went to zero, it would still be difficult to follow a reference from one book to another. In fact, if the cost went to zero, the number of books would increase exponentially, and thus there would be more books than ever between you and the book referred to in a footnote. The great importance of the Web is not that it lets us publish but that it lets us link. And linking does exactly what morality wants us to do: turn toward the world we share and see how it matters to one another.

That certainly does not mean that every link would make Mother Teresa proud. It could easily turn out that the majority of links on the Web point to ads or porn sites. That's why we have spent the past ten years inventing ways to guide one another to the sites that matter to us (including to ads and porn, if that's what you're looking for). The chief method is to say in the text why a reader should click on the link—“You've got to see this hilarious video!”—but we also are busily creating techniques that use the preferences and behaviors of social groups, that analyze patterns of text, that make random stabs in the dark. We are not done innovating—not by a long shot. The potential for understanding—and for letting the world matter to us in new ways—is just too great.

5.

So if saying that links are good is the same as saying that the world is better off with links than without them, and if their goodness resides not just in the quality of the links we're making for one another but in their very structure, in what way are we better off? I think there are two ways.

First, the value of the linked structure of the Web is primarily potential; that is, it is a giant affordance that we may do good or bad with. But it's not potential the way a stick could potentially be used to prop open a car hood. The Web is a potential that we're actively creating and expanding. The potential is the sum of the relationships embodied in links. It is a potential we can traverse any time we're near a browser. It is a potential that can be explored and “mined.” There is nothing “mere” about this potential. It is, so to speak, a real potential, existing and at our fingertips. Fundamentally, it is a potential for seeing how the world matters to others around the spinning ball we share.

Second, we're better off with links because, whether we think about it explicitly or not, every time we click on a link, we take a step away from the selfish solipsism that characterizes our age—or, to be more exact, that characterizes how we talk about our age. We've invested so much in building out the potential of the Web. We've posted tens of billions of pages and created links in numbers that multiply that score. So many of us are so absorbed in this new world that researchers wag their fingers, worried that we're withdrawing from the “real” world.⁸ The Web's reach makes it clearer than ever that the world we share is in fact the entire world, not just our cozy corner of it. The Web's links make it unavoidable that we care about what matters to others, even if we care in the mode of hatred, fear, and ridicule. The world has never seemed so “inter-twingled,” to use Ted Nelson's phrase, and that awareness is a good thing. In fact, it is the very basis and embodiment of morality itself: allowing how our shared world matters to others also matters to us.

Links are good.

NOTES

1. I've purposefully left the bombing of Nagasaki out of this account because it is so much harder even to attempt to justify.

2. I achieve my idiocy in other, subtler ways.

3. Geeks know that link code actually can contain additional metadata, including a phrase like “Hate site hosted by Stormfront.” But there is no agreement about how to encode such data, beyond Google's use of the nofollow tag, which lets a page author indicate that he or she does not want search engines to mistake a link to a page as an endorsement of its worth.

4. A. K. M. Adam uses this empirical argument to make the case for what he calls “differential hermeneutics,” according to which maintaining we're right no longer necessarily entails maintaining that everyone else is wrong. He presents his case in “Integral and Differential Hermeneutics”—a chapter in The Meanings We Choose: Hermeneutical Ethics, Indeterminancy, and the Conflict of Interpretations, ed. Charles Cosgrove (Edinburgh: T and T Clark, 2004), 24–38—which you can read online at http://akma.disseminary.org/06Adam.pdf. It is also the theme of his Faithful Interpretation (Minneapolis: Fortress, 2006).

5. They formulated the Golden Rule in its negative form: don't do to others what you would not have them do to you.

6. The idea that caring is as central to our being as rationality and understanding was the great corrective Martin Heidegger brought to twentieth-century Western philosophy.

7. See Elizabeth Edwards's Saving Graces: Finding Solace and Strength from Friends and Strangers (New York: Broadway, 2006) for a very personal and moving example of this.

8. I use quotes here because there is, of course, only one world.