Thursday, August 28, 2003

Several weeks ago Edward Tufte released his new essay The Cognitive Style of PowerPoint. I received my copy in the mail the other day and spent 20 minutes reading his amusing, but sound, research. I get the feeling that he’ll largely be preaching to the converted with this effort since I know no designers—largely Tufte’s audience—that give presentations using PowerPoint. Mainly, it will fuel the fire of disgruntlement to those audiences imprisoned by PowerPoint’s style.

Before reading, I was curious to know how Tufte would attack PowerPoint for 24 pages—I’ve never known him to show such hostility towards an individual product. But he does an excellent job illustrating the enormous problems with PowerPoint culture—how PP slideshows are for the presenter—a good speaker doesn’t need to read from public notecards. PP bullet lists are making us stupid by heavily dilluting intelligent thought. The resolution of information is lost in the abbreviated format the program forces upon us. “Many true statements are too long to fit on a PP slide.” He occassionally throws in over-the-top jabs comparing Powerpoint users with power-hungry dictators like Stalin and Ceaser, that only delight the anti-PowerPoint reader.

Tufte Versus Nielsen

Not coincidentally, everything Tufte writes about PowerPoint is counter to what Jakob Nielsen has been preaching to us about writing for the web:

Write for scanability. Don’t require users to read long continuous blocks of text; instead, use short paragraphs, subheadings, and bulleted lists.1


[R]eading from computer screens is about 25 percent slower than reading from paper…as a result, people don’t want to read a lot of text from computer screens. Therefore, you should write 50 percent less text—not just 25 percent less—because it’s not only a matter of reading speed but feeling good.2

Sure, Tufte is talking about PowerPoint slides, Nielsen about web pages, but they are at opposite extremes cocerning similar media. When I go to the New York Times web site, I expect the information to be highly detailed and thorough, exactly as it appears in the printed paper. I expect CNN to use many bullet points and subheads because it’s headline news afterall. That, and they generally report the sensationalistic results, not the context in which news stories take place, but that’s a different issue. Each method fits their respective sites’ audience. Nielsen’s tactics are inappropriate because of the wide variety of roles web sites play to their disparate audiences.

User-centered design

Some web sites simply aren’t created for the common good if that notion implies that users be able to find a web page from a Google search and locate the answer to their question (out of context) in five seconds flat so they can leave the site forever. Who creates web sites with that intention? Tufte would also claim to be an advocate of the user (or live audience), and providing background and complete thoughts are often more important than scannability—not that the two are mutually exclusive in the first place—and the full meaning of information cannot be trimmed to a laundry list. Bullet points are good for listing items or displaying linear steps, but not for showing the complex relationship between the items. “The more intense the detail, the greater the clarity and understanding—because meaning and reasoning are contextual,” writes Tufte.

Nielsen shows the following promotional writing example as the “control condition” of a usability experiment:3
Nebraska is filled with internationally recognized attractions that draw large crowds of people every year, without fail. In 1996, some of the most popular places were Fort Robinson State Park (355,000 visitors), Scotts Bluff National Monument (132,166), Arbor Lodge State Historical Park & Museum (100,000), Carhenge (86,598), Stuhr Museum of the Prarie Pioneer (60,002), and Buffalo Bill Ranch State Historical Park (28,446).
The following excerpt was labeled 124% better than the control (with criteria the reader isn’t privy, one of countless examples of dubious methodology in the NN/g), using “all three improvements in writing sytle: concise text, scannable layout, and objective language”:3
In 1996, six of the most-visited places in Nebraska were:
  • Fort Robinson State Park
  • Scotts Bluff National Monument
  • Arbor Lodge State Historical Park & Museum
  • Carhenge
  • Stuhr Museum of the Prarie Pioneer
  • Buffalo Bill Ranch State Historical Park

While the second is undoubtedly more scannable, at what cost are we to dumb-down our data? The loss of data points, the loss of narrative and the loss of intelligence name just a few problems with the “improved” version.

While I’m on the subject, another characteristic of the Nielsen method is that his entire dogma is a reaction to sample audiences’ immediate desires rather than a proactive example of setting high standards. Let’s throw enough shit against the wall to see what sticks the best, tweak it, and throw some more. Imagine writing a song based on audience feedback. (It’s already been done, with results as you’d expect, see Act Two). Nielsen disregards the fact that web sites have their own standards to uphold. If the New York Times followed mass appeal, there would be no more stories of Rwandan elections or goings-on in other non-tourist destinations. Fifteen-thousand word articles about politics in academics would be replaced with short articles of quick facts.

Tufte & The Columbia Accident Investigation Board

This Tuesday’s report by the Columbia Accident Investigation Board has given Tufte’s PP criticism wider exposure—an analysis of a Boeing slide from The Cognitive Style of PowerPoint was excerpted on page 191 of the CAIB report (or page 15 of this PDF), faulting the Boeing presentation with an extremely deficient rate of information exchange. Tufte’s makes the point that had this information been delivered in technical report accompanying the verbal presentation, important data might not have been overlooked. But it was the PowerPoint format—and the ambiguities, fragmentation and noise forced in the inherent PP style—that attributed to engineers not articulating the details. “We need serious methods of communciation for serious problems.”

When a reader argues that it’s the individual PP user at fault for the deficiency, Tufte explains on his site that the product, not the user, is to blame:

Saying that it is a problem with the user rather than the tool blames the victims of PP (the audience, the content, and the user)…This evidence [from examining thousands of PP presentations] points to inherent defects, unless one advances the entertaining hypothesis that nearly all PP users are stupid and that nearly all users of other methods are not. PP’s inherent defects are a much more likely explanation. That explanation also has a direct practical prescription—abandon PP—rather than asking millions of PP users to learn tricks in a vain attempt to undo inherent problems in a slideware computer program.

As for Apple’s Keynote presentation software, he makes no mention of it in his essay, but does say this on his site:

Alas the Keynote examples are as data-thin as PowerPoint. Only a few data points, no multivariate examples. Both Keynote and PP are tinker toyish.

An (ironically) abbreviated version of The Cognitive Style of PowerPoint appears in the September issue of Wired magazine as PowerPoint is Evil, juxtaposed to David Byrne’s essay Learning to Love PowerPoint that promotes Byrne’s new PowerPoint art DVD. But to get Tufte’s full essay, I recommend sending $7 to Graphic Press.

  • 1Jakob Nielsen, Designing Web Usability (New Riders, 2000), 101.
  • 2pp. 101-102.
  • 3p. 105.

Other articles on Tufte and Powerpoint:


by Keith » Aug 28, 2003 11:29 PM

Great piece. A quick comment. To your point about writing for the Web, if you have a chance and haven't yet, check out what Gerry McGovern has to say. He falls somewhere in the middle of Tufte and Nielsen (neither of whom I'd classify as a writer first) and has quite a bit more insight into Web writing than either. His book, Content Critical, is quite good.


by michael » Sep 2, 2003 8:42 AM

More PPT and Tufte articles at powerpointless?.


by Paul Souders » Sep 4, 2003 2:11 PM

Oh I have way too much to say about this topic. PowerPoint is the major mode of communication with our clients.

Our usual use for the thing is like this: the accountsfolks email a PPT to a client, then we "walk them through it" on a phone call. If I were a client this would be incredibly insulting. But our top client (a big hardware company) communicates internally almost entirely by Powerpoint, using the same MO, so they must be conditioned to it. We frequently create PPTs for our client contacts to give to THEIR colleagues, and THEY use the email/phone call attack pattern.

Nondesigners love to use Powerpoint for, ahem, design. Clients, writers, accountspeople...whenever they have to draw a spider diagram or wireframe or just a damn box they create a PPT with a single lonely slide. There's a market in there somewhere I'm sure.

The other minor function to which it seems to be employed is as an outliner which I have to admit it does pretty well. The bullets nest themselves. But doesn't Word do outlines? I dunno, I've used Word for 11 years and never understood the outliner.

I'd say fewer than 1 in 20 PPTs coming through this shop are ever used in an actual stand-up presentation.

I have no idea what point I'm trying to make here. Except I really hate seeing nested bullets. Yeah, that sucks.


by Scott Steffens » Sep 4, 2003 5:28 PM

A written report is preferable in nearly every case. Except when a graphic of some kind is the topic of a presentation, Word does a better job of explaining.

Even bullet lists are better in a word processor because you can fit more information in the bullets to present complete thoughts, make lists longer and contextual instead of spanning slides or arbitrarily shortening them to fit.

But complete failure happens when someone can't attend a meeting and the PP deck is e-mailed and must stand on its own.


by Michael McWatters » Sep 20, 2003 3:36 PM

Having worked in a well-respected consultancy for a while, I found another reason to despise the way PowerPoint was being used.

Our clients were often overwhelmed by 150-250 screen PowerPoint presentations that often took 2 or 3 hours to get through. I argued, unsuccessfully, that these presentations could be far more condensed, with a detailed paper document given out as a takeaway or leave-behind.

Simply put, some consultants would spend months pouring their research into these bloated PowerPoint decks, and then practically recite them verbatim to the client. Not surprisingly, members of the consultancy team and client team would fall asleep, groan, sigh, or lose interest.

In the end, however, I began to realize that the PowerPoint deck was considered THE project deliverable during the strategy phase, so the fact that it could look "pretty" became one of the primary reasons it was chosen over a detailed Word document.


by Joo Khim TAN » Sep 22, 2003 2:21 AM

Very interesting commentary. Now, I am sufficiently interested to want to read the Powerpoint is Evil article (whose title seemed rather frivolous) or the actual paper in full. Just wait till I can find the time.

For the moment, just a few comments:

1. I don't agree with the "dubious methodology" example. The same text can be re-written as follows , if the details are deemed important enough:

Nebraska is filled with internationally recognized attractions that draw large crowds of people every year, without fail. In 1996, six most-visited places in Nebraska were:
- Fort Robinson State Park (355,000 visitors)
- Scotts Bluff National Monument (132,166)
- Arbor Lodge State Historical Park & Museum (100,000)
- Carhenge (86,598)
- Stuhr Museum of the Prarie Pioneer (60,002)
- Buffalo Bill Ranch State Historical Park (28,446)

Nothing is lost. But the text is now more scannable than the original.

2. I don't see the choice of presentation style as a case of Tufte vs. Nielsen. I would think that there's a place of either, depending on the audience, context and content.

Nevertheless, thank you for a most interesting commentary on Tufte's paper.


by miglia » Oct 1, 2003 12:53 AM

I've found myself wondering what it is exactly that makes PPT evil. Certainly it is dangerous: a graphic communications tool in the hands of people poorly trained in graphical communication is a bad thing. As Tufte points out, hierarchical outlines can be used to lend a spurious authority to banal or misleading statements and imply non-existent chains of inference and conclusion. But this, I think, is not enough to make PPT truly evil. For a long time I wondered what I was missing, until I came across this:

Leverage your existing presentations so you don’t have to start from scratch. You can import just about any file type into Keynote - including PowerPoint, PDF and AppleWorks presentations - and then enhance with themes. You can paste data from Excel documents into your Keynote charts and tables. Keynote lets you export presentations to PowerPoint, QuickTime or PDF.

Here ... and I realised that Chomsky had answered the question over a generation ago.

PPT, surely, has as its antecedents the blackboard, the flip chart and the ohp. Even used amateurishly, all of these media are effectively deployed in communication. Thinking back to my schooldays, I was always worried about teachers who flourished OHPs rather than wrote on the board, for some obscure reason, but they never struck the terror into me that a session of PPTs can. Why is this? And why did ohps make me more nervous than blackboards?

In the 1970s Chomsky noted that television was destroying political discourse. He realised that, in fact, discourse was stopping, as television demanded immediacy, and is not well suited to the delivery of lectures, encouraging a style of discourse now known as the "soundbite". At first, "soundbites" were the distillation of more complex arguments - and this was the point of Chomsky's objection: that complex political debate was being "dumbed down" into a soundbite for television's consumption.

This was the effect of television itself--as McLuhan spotted, the medium is the message--but the political classes soon got with the medium and rather than "dumb down" the argument to get to the soundbite, dropped the argument entirely to produce just the soundbite. By the 1980s, politics had become merely soundbite packaging: Consider, since when did "tough on crime, tough on the causes of crime" actually substitute for a policy on criminal justice?

Although politics has always been about sloganeering--wrapping a complex idea into a memorable phrase like "votes for women", "peace in our time", "liberty, equality, fraternity"--there used to be complex political ideas behind the slogans. Nowadays, political parties don't have policies as such, they instead craft soundbites to appeal to target swing voter groups. The party that does this best gets elected.

There are no longer any big ideas in politics not because all the big idea battles have been won, but because there are not anymore big ideas at all - and PPT has helped this happen to the presentation of complex information.

In the past, the notes on the blackboard represented a summation. The teacher wasn't writing all there was to know on the subject - that existed in books, papers, pictures, documents, films, and other archives. The teacher merely presented a synthetic overview of the corpus relevant to the lesson at hand.

The teacher was able to do this (if they were a good teacher) because they had some mastery of that corpus. The notes on the board were ephemeral, epiphenomena of the narrative the teacher's master caused him/her to weave around the source material. On reflection, this is why I got nervous about OHPs.

OHPs were more difficult to produce, and were produced in advance of the lesson. The teacher became preoccupied with the presentation of the OHPs, making sure they were laid out clearly and legible from the back of the class, as they would be unable to effect significant changes on the fly. They would have to prejudge very accurately the length of their talk, and the level of engagement of their audience. They would, in short, have come to see the production of the OHPs as the end in itself, rather than the summative mastery of the subject matter.

PPTs, too, has become an end in itself. PPTs don't summarise more complex corpora, they are the sole embodiment of a piece of thinking, information or ideas. The are lavishly prepared: my anecdotal impression is that for every hour a PPT is worked on, 40 minutes are on looknfeel, and 20 minutes are on content.

As more and more visual tools are loaded into presentation software, as with Keynote, more and more time is spent on the looknfeel. This is what makes PPT evil: it is the primary medium for the expression of ideas in business, and, increasingly, education.

PPT is no longer an ephemeral medium, but a medium of record - so what we record is executive summaries and bullet-points. Not only are complex ideas no longer explored --if they won't fit on a slide, there's no place for them--but people are becoming increasingly ignorant of complex ideas: All thought has become slogans.

Is there hope? Very little, I fear. But I say this - delete your PPT slides after presenting them. Promise yourself that you will always treat them as ephemeral, that your primary sources will be elsewhere, in greater depth, and with more detail, and you may yet be saved.


by Daniel Brashler » Dec 19, 2003 8:22 AM

At one time, I think that "public speaking" was regarded quite highly as an art, and valued as a skill in any workplace or forum. Doing it well required a synthesis of intellect, aesthetic and passion that not only communicated information, but truly motivated action as well as differenting people from one another as leaders and followers.

Sadly, though faced with a tremendous requirement for such skills in so many areas, our society has produced only half-hearted and cursory attempts to teach the skills needed for effective communication. It's thought that reading and writing are the whole of the issue, and that's all we've been able to tackle. I fear that on one hand, serious research on effective communication -- research that might show what things beyond putting sentances together with decent grammar -- is often smiled-and-nodded-at as "simply common sense" or somewhat maligned as "handy tools for hucksters." The fact is that our society's attempts to use this work to it's benefit are paltry. The classic mode of teaching these things, at least at a school-age level, is the old "Speech and Debate Team". Is there a group of people more widely derided as geeks and losers?

The fact is that skillful communicators use whatever tools are available, and know, when they prepare to communicate, what they want to do with them to make them the most effective. If they want to prepare a slide with three bullets, PP fills the bill. If they need details read, they make a white paper with Word.
Perfunctory communicators simply assume that their preparation and use of available are adequate regardless of how they're used.The problem with PowerPoint or any slide-maker program is the same as with any tool.

Buying PowerPoint doesn't make me an effective communicator any more than buying a wrench makes me a master mechanic, no matter how much advertizing money is spent to lead me to believe it, or how many wizards are programmed into it. As well, even while the art of effective communication might be teachable, giving people methods for understanding how to convey information, the basic skill of being able to move one's own information into one's own audience will never be something that you can buy in a box from someone else who knows neither of the two.



RSS: .91 / 1.0 / 2.0