joho logo

February 4, 2008

Receive email issues. Free!



Is the Web different? Is the Web just the next medium in our history of media, or is it a spiritual transformation, the great hope, blah-di-blah-di-blah?

Fairness and scarcity: In a world of abundance, fairness is so 1990s. 

The next future of HTML:  The draft of the next version of HTML manages a surprisingly fine balance between the needs of humans and the needs of our computer overlords.

Bogus Contest: Tech clichés


Moi moi moi ... now with added Facebook!

I assume subscribers to this newsletter know that I blog every freaking day, but just in case, here's the link:

I also twitter:

I also maintain a blog about Everything Is Misc at ( works also.)

If you want to chart my receding hairline, I post other people's video interviews with me at

Pretty soon now I'm going to update my list of publications, etc. Right now, it's seriously out of date.

I've even set up a Facebook group for Joho readers, just because that's what the cool kids are doing. It's called — surprise! — Joho. You have to join Facebook to avail yourself of it, which means you have Yet Another inbox to check. But once there, you'll be able to do something or other. It beats me. Let me know if you figure it out.

Teaching again

I'm teaching a college course again, for the first time since 1986. This doesn't make me anxious at all. After all, teaching is just like riding a bicycle: Even after all these years, you just get back on, slam into mail boxes, and look way worse in spandex.

The course is at Harvard Law, which doesn't ratchet up my anxiety even seven or eight orders of magnitude. Fortunately, my co-teacher is John Palfrey, who is (a) wise beyond his or anyone's years, (b) one of the best teachers I've ever seen, (c) the sweetest man on the planet. Our topic is: Is the Web different from what came before it, and what effect does and should that have on law and policy?

You can read the class blog here. There's a syllabus posted there, too.

dividing line
Is the Web Different?

The question "Is the Web different?" is actually not so much a question as a shibboleth in the original sense: The answer determines which tribe you're in.

The Web utopians point to the ways in which the Web has changed some of the basic assumptions about how we live together, removing old obstacles and enabling shiny new possibilities.

The Web dystopians agree that the Web is having a major effect on our lives. They, however, think that effect is detrimental.

The Web realists say the Web hasn't had nearly as much effect as the utopians and dystopians proclaim. The Web carries with it certain possibilities and limitations, but (the realists say) not many more than other major communications medium.

Each of these is a political position: They imply normative beliefs, and they lead their holders to certain types of behaviors and actions:

The utopians want the Web to have wide effects as quickly as possible. They therefore favor connecting as many people as possible and maintaining the Web as an open, public space.

The dystopians want to curb the excesses of the Web, or prepare us to deal with those excesses.

The realists want to curb the excesses of the utopians who, they think, are feeding unrealistic expectations.

Simply the act of holding the position is itself a political action for all three groups:

The utopians think that by holding out a vision of what will or might be, they are affecting the direction of the present.

The dystopians are sounding a call to action, even if some dystopians think that we are doomed to suffer under the Web's increasing hegemony.

The realists may not view their position as political because it is — they believe — based merely on a clear-eyed, non-politicized view of the world. But this is itself a political decision that leans toward supporting the status quo because what-is is more knowable than what might be.

So, which of the three positions — or some variant — is right? Is the Web different in a way that matters?

The obvious answer to the question "Which one is right?" is: Time will tell.

Unfortunately, time papers over all wounds. Our values change, so our evaluations of change shift over time. The extraordinary becomes ordinary with extraordinary rapidity and insinuates itself into memory, undercutting the reliability of our judgments about the magnitude of change. So, time will not tell.

Nor is this a simple fact-based issue. Realists would like it to be, but that's what makes them realists. Consider this hypothetical exchange:

Realist: You say that the Web will transform politics. But politics is as it ever was.

Utopian: Just wait.

This is, indeed, one of the two basic blocking tactics used by Web utopians: The changes are so important that they will take a while to arrive, and the changes are so fundamental that we aren't always even aware of them. Here's an example of the second tactic at work:

Realist: You say that the Web will transform business, but business is as it ever was.

Utopian: Not at all! For example, email has transformed meetings, but we're so used to the change that we don't even recognize it.

To this, the Web realist has a number of responses: Denying that the changes are real, that they are important, or that they are due to the Web.

When a dystopian points to a bad effect of the Web, the utopian denies the truth of the value claim, its inevitability, or its importance:

Dystopian: The Web has made pornography available to every schoolchild!

Utopian: It is the responsibility of parents to make sure their kids are using child-safe filters. Besides, viewing pornography may weaken our unhealthy anti-sexual attitudes. Besides, greater access to porn is just one effect of the Web; it's brought greater access to literature, art, science...

The realist wants to bring the argument squarely within the realm of facts. Facts can, of course, resolve some disputes. But facts are unlikely to settle the overall question of the Web's difference because the utopians, dystopians and realists are probably operating from different views of history, and the framing of history also frames facts.

Many utopians think the Web has uncanny power because they are McLuhanites who think media transform institutions and even consciousness. The McLuhanites' belief in the shaping power of media leads them to a rhetoric of "not only": Not only did the printing press enable the spread of literacy, it led to our reliance on experts. The next McLuhanite up says, "Not only did it lead to experts, it actually changed the shape of knowledge." Web utopians engage in the same rhetorical one-upmanship.

Many Web dystopians share the utopians' disruptive view of the Web, although they are struck more by the facts with negative values.

Many Web realists think change happens far more incrementally. They feel the inertial weight of existing institutions and social structures. Nothing as trivial as HTML will change the fact that most of the world is in poverty and that corrupt corporations are firmly in control.

These positions about how history works cannot be defended by looking at history, for they determine how history is to be read. For example, did the Howard Dean campaign in 2004 show that the Web is profoundly altering politics, that the Web has had little effect on politics, or that the Web is further degrading politics? All three positions are defensible because historical events such as presidential campaigns are carried along social wavefronts of unfathomable complexity. Did Dean get as far as he did because of the Web or because of the media? Did his campaign fail because the Web created a bubble of self-involvement, because the Web ultimately did not get people out to vote, or because he was a quirky candidate who, without the Web, wouldn't have been noticed outside of his home state of Vermont?

To make matters yet more complex, holders of these three positions are not merely uttering descriptive statements. Frequently, they speak in order to have a political effect:

Utopians want to excite us about the future possibilities because they want policies that will keep the Internet an open field for bottom-up innovation. 

Dystopians want to warn us of the dangers of the Web so we can create policies and practices that will mitigate those dangers. 

Realists want to clear away false promises so we can focus on what really needs to be done. Also, they'd like the blowhard utopians to just shut up for a while.

Arguments that have different aims and are based on differing views of how history works and of the nature of the interactions between the material and social realms are not settled by facts. In fact, they're not settled. Ever. Even after the changes happen, these three temperaments and cognitive sets will debate why the changes happened, how significant they were, and whether they were good, bad or indifferent.

Time won't tell.

Unfortunately, we can't afford to wait for time not to tell us. "Is the Web different?" is an urgent question. Decisions depend on our answer.

For example, if the Web utopians are right — if the Web is transformative in an overall positive way — then it's thus morally incumbent upon us to provide widespread access to as much of the world as is possible, focusing on the disadvantaged. If the Web dystopians are right, we need to put in place whatever safeguards we can. If the realists are right, then we ought to make tactical adjustments but ignore the hyperventilations of the utopians and dystopians.

Then there are the more localized decisions. If the Web is transforming business, for better or for worse, then businesses need to alter their strategic plans. If the Web is merely one more way information travels, then businesses should be looking only at tactical responses. Likewise for every other institution that deals with information, including government, media, science, and education.

So, we need to decide.

But there is no way to decide.

Fortunately, this is not the first time we humans have been in this position. In fact, it is characteristic of politics overall. Who's right, the liberals, the conservatives, or neither? Because such a question can't be answered to the satisfaction of all the parties involved, we come up with political means for resolving issues. For politics to work in helping us to decide what to do about and with the Web, we need all three positions plus the incalculable variants represented.

Together we'll settle the future's hash.

But I don't want to leave it at that happy, liberal conclusion because it is, I believe, incomplete. The fuller statement of the conclusion should include: It is vital to have realists in the discussion, but they are essentially wrong.

I am using the word "essentially" carefully here. Web realists are often right in their particular arguments, demurrals and corrections, and the utopians and dystopians are often wrong in their predictions, readings, and even facts. That matters. Yet, the essence of the utopian and dystopian view is that the Web is truly different. About that they are right.

Why? I am enough of a McLuhanite to believe that media do not simply transmit messages. The means by which we communicate has a deep, profound and even fundamental effect on how we understand ourselves and how we associate with one another. Yes, the medium is the message.

If that's the case (and notice I am not giving any further argument for it), then there are good reasons to think that the Web as a medium is likely to be as disruptive as other media that have had profound effects on culture. Perhaps the best comparison is to the effect Gutenberg's invention has had on the West. Access to printed books gave many more people access to knowledge, changed the economics of knowledge, undermined institutions that were premised on knowledge being scarce and difficult to find, altered the nature and role of expertise, and established the idea that knowledge is capable of being chunked into stable topics. These in turn affected our ideas about what it means to be a human and to be human together. But these are exactly the domains within which the Web is bringing change. Indeed, it is altering not just the content of knowledge but our sense of how ideas go together, for the Web is first and foremost about connections.

Clearly, there is much more to say about this, and much has already been said. But that is the general shape of one Web utopian argument.

It can, of course, be challenged. It should be challenged, both in its outline and in its particulars. Here Web realists have a vital role to play. But at the highest level of abstraction, these three positions are not truly arguable. Each is an expression of an attitude towards the future, and the future is that which does not yet exist. None of these three positions truly knows what the future holds if only because the prevalence of these positions itself shapes the unknown future.

And that is a reason to join the utopian tribe, or at least to acknowledge the special value it brings to the conversation. Innovation requires the realism that keeps us from wasting time on the impossible. But some of the most radical innovation requires ignoring one's deep-bred confidence about what is possible. This is especially true within the social realm where the limits on new ways to associate are almost always transgressible simply by changing how we think about ourselves. We thus need utopians to invent the impossible future.

And we need lots and lots of them. There is so much to invent, and the new forms of association that emerge often only succeed if there are enough people to embrace them.

Web realists perform the vital function of keeping us from running down dead ends longer than we need to, and from getting into feedback loops that distort the innovation process. For those services, we should thank and encourage the realists. But we should also recognize that beyond the particulars, they are essentially wrong.

The contention among dystopians, realists and utopians is is a struggle among the past, the present and the future. The present is always right about itself but — in times of disruption — essentially wrong about the future. That's why we need to flood the field with utopians so we can be right often enough that we build the best future we can.

It is, of course, simply an accident that this defense of Web utopianism comes from someone who is personally a Web utopian. Absolutely coincidental.

Uh huh.

dividing line
Fairness and scarcity

Time-Warner Cable (TWC) recently acknowledged that it's going to test a billing system that will move Internet access closer to the  cellphone model: Those in the test will subscribe to a tier of service that buys them a certain number of bytes (like buying a package that gives you 500 minutes of cellphone time), and if they go over their allotment, they'll pay per byte.

This certainly seems fair. And it's better than other, threatened ways of limiting the amount of network traffic. But, in my opinion, it's ultimately a bad way to go. Being fair is not enough. In fact, sometimes what's fair is wrong precisely because it's fair.

Oooh! A seeming paradox! One of the top three rhetorical forms for essays!

TWC's proposal is a welcome relief from the Internet carriers' arguments against Net neutrality. (See box below.) It lets users decide whether they want to spend their Internet allotment on, say, lots and lots of email or a few high-definition movies. Users could decide that doing VOIP, which burns through bits with some rapidity, is worth it even if that means they can't do all the Facebooking they might want to do. Letting users decide is way better than letting the carriers decide that we all want Hollywood movie packets to shoulder aside World of Warcraft packets, YouTube packets, or Nigerian spam packets. 

Nevertheless, I don't like the TWC proposal. It's fair but it's a good example of where fairness can get in the way.

I've written about the fairness argument before, when arguing against the tit-for-tat view of "intellectual property" that says a creator ought to be paid every time her audience gets any value from her work. That's fair, if fairness means an equivalence of value in an exchange.1 But it leads to a worse world. And that's why I don't like the TWC proposal. In both cases, we reduce our cultural context to a transaction. We have to think and calculate before we engage with one another and with what we create for one another. 

It doesn't take much friction to disrupt a social or cultural ecology. Think about how badly having to watch what you say disrupts your relationship with your boss or your prospective in-laws. Having to watch what you read or hear is just as disruptive. That is precisely what making fairness our highest value will do to culture, whether that fairness is the basis of our increasing copyright totalitarianism or TWC's pay-per-packet scheme. Fairness will change the Internet from a world into a resource. Abundance doesn't work its transformative magic if you have to justify your every use of it. That brings scarcity-thinking to the  abundance table.

Fairness in fact operates best in times of scarcity. When you're trapped in a subway tunnel with only enough food for three days, and the zombies are scratching at the rubble, fairness is a good way to divide up the remaining saltines, although you'll have to argue whether it's fair to give out five per person, or to give the big folks more than the wee folks, or maybe to let market forces determine the outcome. Nevertheless, you'd be right to use fairness as your guide.

In times of abundance, tit-for-tat fairness plays a different role. Rather than being a positive principle, unfairness becomes a bottom limit, a minimum standard. That a billing scheme is fair doesn't tell us that it's the right one, but if it were unfair — half off for Aryans! — that would tell us that it's the wrong one.

Utopianism is frequently a better guide than fairness in such cases. If we could pay for Internet content any way we wanted, what would wring the maximum social, cultural, political and economic value out of this new infrastructure? I think the answer to that question isn't all that controversial: Everyone would have all they could use, everywhere they are, without price much inhibiting their participation...while still providing sufficient incentives so that creators will continue to create. How we get there is, of course, subject to tons of debate. But our aim should be to make the Net so abundant that fairness is irrelevant.

Why I'm not neutral about Net neutrality

Net neutrality, as I understand it, is the fundamental architectural principle of the Internet that says that all packets will be treated equally. It thus prevents Internet carriers from discriminating against packets based on their origin or application type. 

The Internet carriers' strongest argument2 against Net neutrality goes roughly as follows:

1. Network traffic is overwhelming capacity.

2. Therefore, we have to limit network traffic somehow.

3. Some types of traffic are more time-sensitive than others: You don't care if your email arrives one second later, but you do care if your Internet-based telephone call or movie jitters by a second.

4. Therefore, carriers ought to be allowed to "shape" net traffic by delaying non-time-sensitive packets, and hurrying the time-sensitive ones.

This argument fails (imo) because: (i) It assumes that, if indeed premise #1 is true, the carriers, who are in the business of selling us content, are the best ones to decide which content to deliver fast. In fact, there are better alternatives, such as letting each user decide that, say, she wants World of Warcraft bits to get priority over CD-quality phone calls, or that she'd prefer to get jitter-free YouTubes but never watches Hollywood movies. (ii) This argument gives carriers a financial incentive to keep Net connectivity scarce.

(Just in case you're keeping score: Obama supports Net Neutrality, Hillary doesn't talk about it, and as far as I can tell, none of the Republicans do.)

1Note that I am not using fairness in the way most prevalent among philosophers since 1971 when John Rawls published A Theory of Justice. I mean it in the simple sense that a transaction is fair if what's given is roughly of the same value as what's taken.

2The carriers frequently add a second argument: Some sites generate more traffic than others, so it is only fair that they pay more if they want their packets delivered speedily. But: Those sites already pay for the packets they're producing, and letting some sites respond more speedily by paying off the carriers works against small sites and thus hurts innovativeness. 

dividing line

The next future of HTML

Remember back before HTML, when SGML was battling to be the way software expressed a document and its structure? SGML was precise and kept every hair in place, while HTML was ok with some ambiguity and hadn't showered in a couple of days. With the release of a draft of HTML 5, we see that the battle is not over. Far from it.

SGML lets you specify all the parts of a document and how they go together: A cookbook might have elements such as recipe, list_of_ingredients, instructions, photo and notes, and you might set up rules such as: "Every recipe must have a list_of_ingredients and instructions." You could do this in infinite detail. In fact, SGML was such a fine standard that entire industries came to a standstill as they tried to perfect the  structures required for complex document sets such as aircraft repair manuals and telecommunications equipment specification sheets.

Then along came HTML, which is the SGML specification of the elements a Web page can or should have. HTML said that a Web page can have six different types of headings (H1 ... H6), paragraphs (p), links (A), etc. It was so simple and incomplete the SGML-ers generally referred to it as "brain dead." But HTML ruled because it was so simple, and enabled us to make clickable links that brought to each individual document the value of the web in which it was embedded.

HTML succeeded also because Web browsers had an incentive to forgive its trespasses. SGML systems were generally installed in controlled, disciplined environments where you could insist that your writers use no <p> without a corresponding </p>. But HTML was taken up by undisciplined amateurs who just wanted to type 'n' post. They didn't want to run a spellchecker much less an arcane syntax validator. What benefit does the person posting directions to her house or instructions for setting a Casio watch get from worrying about syntax? So, the browsers forgave just about all mistakes. Competition assured this. If you were making a new browser, you wanted it to be able to read every page your competitors could, and more.

The browsers then fought for dominance in part by coming up with their own structural elements, hoping that people would create pages using them, so that their browser would display something that other browsers did not. Some were useful. Some were the blink tag.

Such tags drove the standards folks crazy. Take the blink tag. SGML-style standards folks hated it because: (1) It's ugly. (2) It came from a single vendor (Netscape Navigator, way back when) and thus was not uniformly accepted. (3) The blink tag expresses how information is displayed, not anything about the document's structure. For that same reason, SGML folks don't like the font tag. 

SGML's preference for structure over format (or "presentation," as they way) is simple, powerful, and annoying.  For example, SGML would let you note that this is a headline, that is a by-line, and that other thing is body text, and it could let you specify that every article has to begin with a headline, optionally have a by-line, and always be followed by body text. But SGML was not designed to let you say that headlines are in 24pt type and centered while by-lines are in 10pt type and italicized. That sort of formatting information was to be kept in a separate file that defined the stylistic properties of the various elements. Put in new style definitions, and suddenly your newspaper goes in appearance from NY Times to NY Post, while maintaining the same structure — the lead story is the same, the articles have by-lines, etc. Separating structural and formatting information is an amazingly powerful, even liberating idea.

The problem is that most of us aren't standards folks and we don't write by dividing documents into structure and format. Certainly when we read we don't: We use format as a guide to structure. And we know from 20+ years of word processing wars that we don't write by separating the two. Back in the late 1980's, WordPerfect was kicking Microsoft Word's skinny butt by letting you create, say, a title by hitting the "center" key and the "bold" key. Word, on the other hand, encouraged you to define the line as a "title," and then give formatting properties (centered and bold) to "title" elements, in quite an SGMLy way. It turned out that few people actually wanted to do that, so they instead "misused" Word by using it exactly how they used WordPerfect. You still see this behavior in Word users who separate paragraphs by hitting the Enter key twice. Foolish mortals! The proper SGMLy method is to define your paragraph element as having a certain bottom margin and only hitting the Enter key once. Otherwise you are creating a structural element (a paragraph) simply to accomplish a formatting aim (putting space between the paragraphs), which is a violation that can cost you your structured document driving license. In fact, you should probably just sign in to your nearest SGML rehab center.

So, now the HTML standards folks are ready for us to take the next big step forward. We are currently in official version 4 of HTML, first published in December 1997. Despite the fact that we should beware any standard ten years in the making, HTML 5 it attempts in a sensible way to let us have our structured cake and eat its format too.

HTML 4 lets us work the way most of us want to: We use the basic HTML elements, but we format them in the WordPerfect way. So, if we don't like the default format of an <H1>, we'll put in some code like this:

<h1><font face="Arial" color="red"><u>Howdy!</u></font></h1>

Certainly, these days many of us would instead do the right thing, which is to have our HTML reference a format file (CSS) that defines H1 as being red, Arial, and underlined. Even so, if we want an exception — we want this particular H1 to be green or underlined — we won't bother creating a special class of H1 via CSS. We'll just stick in a <font> tag that colors it and a <u> tag that underlines it (or we'll insert a snippet of CSS style info right into our HTML). The SGML folks may snicker, but it's just not worth it to us to open up a CSS editor and make the change.

Although HTML 5 doesn't like the <font> element, it recognizes something that SGML long struggled against: Not only do computers and humans read documents differently, each has its place. So, HTML 5 introduces a distinction between HTML done for and by "user-agents" (browsers and other programs that handle pages) and authors. And authors are given concessions. Sure, it'd be better to separate all the formatting info from all the structural and content info, but it's not gonna happen so long as humans are in control. So, the <font> tag survives. Sort of. One of the W3C docs puts it this way: the font tag "is allowed when inserted by a WYSIWYG editor due to limitations in the state of the art in user interface for these editors."

Other formatting tags escape unscathed, although they are redefined in non-formatting ways, often quite awkwardly. For example, the <b> element now "represents a span of text to be stylistically offset from the normal prose without conveying any extra importance, such as key words in a document abstract, product names in a review, or other spans of text whose typical typographic presentation is emboldened." In other words, it's defined as an element worth bolding and that is typically bolded, but not as an element that is bolded. I'm sure that distinction makes someone happy. In any case, HTML 5 supports <b> pretty much as it always has. And the new <m> element indicates marked or highlighted text; while that does not dictate how the element should be highlighted — yellow overlay? red box? — it acknowledges that some document structures are inextricably tied to their display.

The underline and strikethrough tags (<u>,<s>) are discontinued in HTML 5 because they are "purely presentational," although they don't seem any more or less structural than bold and italic. There must have been some fun debates about these on the HTML 5 mailing list.

At the same time, HTML 5 introduces some obvious structural elements lacking in HTML 4. Most important, the <section> tag will tell browsers and other apps that you mean to divvy up your page in a structural way.  Before that, people usd the <div> tag to indicate sections, but you can also use the <div> to mark any stretch of a document, not just its sections. The <section>

tag comes with a meaning already set, marking a structural element of documents common enough that it deserves to be a built-in part of the semantics of Web pages.

Similarly, the <figure> tag lets documents express a common structural relationship: This graphic (or video or whatever) goes with this caption (or whatever). And an <aside> is of peripheral interest. And an <article> is an "independent piece of content" such as a blog post or a newspaper article. All of these new HTML 5 elements are structures common enough that they indeed deserve their own tags. Trying to come up with HTML 5 elements much more specific than this would have driven it into the deep weeds that swallowed so many SGML efforts.

This is by no means a full representation of all that's in HTML 5. Nor am I a competent reviewer. I am impressed, however, that the upgraded standard leaves wriggle room for imperfect humans to mix structure and formatting instructions, rather than insisting that we always structurally separate structure and format the way our computers would like us to. HTML 5 favors the computer view, but leaves room for us silly humans.

dividing line

Bogus Contest: Net clichés

These days, instead of saying "If you look up 'miserable failure' in the dictionary, there's a picture of George Bush there," you'd more likely say, "If you google 'miserable failure,' George Bush is the first return."

Can we come up with more clichés transposed to the world of tech? For example:

The more things are upgraded, the more they stay the same

A watched IPO never boils

It takes two to flame

A woman needs a man the way a fish needs a C compiler

There's more than one way to skin a Firefox

When the net nanny's away, the mice will play

Your turn! (To enter — which is, remember, functionally the same as not entering — post your updated clichés here

Editorial Lint

JOHO is a free, independent newsletter written and produced by David Weinberger. If you write him with corrections or criticisms, it will probably turn out to have been your fault. To unsubscribe, send an email to [email protected] with "unsubscribe" in the subject line. If you have more than one email address, you must send the unsubscribe request from the email address you want unsubscribed. In case of difficulty, let me know: [email protected]

There's more information about subscribing, changing your address, etc., at In case of confusion, you can always send mail to me at [email protected]. There is no need for harshness or recriminations. Sometimes things just don't work out between people. .

Dr. Weinberger is represented by a fiercely aggressive legal team who responds to any provocation with massive litigatory procedures. This notice constitutes fair warning

The Journal of the Hyperlinked Organization is a publication of Evident Marketing, Inc. "The Hyperlinked Organization" is trademarked by Open Text Corp. For information about trademarks owned by Evident Marketing, Inc., please see our Preemptive Trademarks™™ page at

Creative Commons License
This work by David Weinberger is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.