All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Envisioning Information
Visual Explanations
Beautiful Evidence
Paper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
All 4 clothbound books, autographed by the author $150
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer:
Visual and Statistical Thinking $2
The Cognitive Style of Powerpoint $2
Seeing Around + Feynman Diagrams $2
Data Analysis for Politics and Policy $2
catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Tampa, March 1
Boston, March 14, 15, 16
Oakland, April 20
San Jose, April 21
Palo Alto, April 24
San Francisco, April 25, 26
New edition of "The Cognitive Style of PowerPoint"

The first printing of "The Cognitive Style of PowerPoint" has sold out. I have now published a new edition with 4 more pages. The additional material contains the results of the Columbia Accident Investigation Board, a delicious quote from Richard Feynman about bullet lists, and responses to sometimes asked questions (Isn't PP just a tool? Do I have to use PP because all my colleagues do? What are PP-free ways to give a talk? What about that NASA slide with the Earth at the center of universe?).

Little of the additional material will be new to readers to this board. I am most grateful for your comments and questions.

-- Edward Tufte

The copies dated "May 2003" in the copyright notice represent the "first edition," a matter probably only of interest to bibliophiles. The new 28-page version is dated "September 2003."

The internet has changed the course and pace of paper publishing, at least in the case of my essay. About 25% of "Cognitive Style" first appeared on this board (the cancer data analysis, the Columbia slide, the magical number 7) over a period of 6 months, then came the broadside booklet, then came new material developed on this board, then a revised booklet . . . .

In the long-run, I much prefer the extended, coherent narrative and evidence of the booklet, but it has certainly been helpful to publish the short case studies on this board. The internet was ideal for the Columbia material, getting that analysis out instantly so that it could be part of public discussion about the accident and part of the report of the Columbia Accident Investigation Board.

But one way or another, if it isn't published (on the internet or on paper), it can't make a difference.

-- Edward Tufte

Shown are character counts, not word counts.

Instead of the tinker-toy default typography of IE and try Netscape 7.0, with preference fonts set as follows: proportional = serif 11, serif = Times, sans serif = Helvetica, monospace = Courier 11, and minimum font size = 9, all of which are probably the default fonts in Netscape 7.0 at least on the Mac G4, Cinema monitor. (Similar results were obtained on a G4 Powerbook using Netscape.) These screen fonts are perfectly readable; they are my default fonts, used all the time.

Netscape images were then printed out on 8.5 by 11 inch paper for the character counts. Again the printed fonts are perfectly readable, although a bit small.

Is there a way to avoid the overly large IE fonts (except by setting the view at text zoom of 75% for a given document, a preference that does not last beyond that document)? Somehow on the Powerbook, the IE fonts look reasonable (where IE is my main browser) but not on the Cinema monitor (where Netscape is my main browser).

-- Edward Tufte

I have the May edition. Do you offer any resource to upgrade, such as the four new pages in .pdf to add as an insert, or do I consider buying again? I perfectly understand that an author's natural response is to answer: "Buy another!".

For shaping the text content, I find a useful resource in Martin Cutts; The Quick Reference Plain English Guide - How to write clearly and communicate better, Oxford University Press, 1999. It includes something like the Minto Pyramid Principle (mentioned above), plus a dozen or more other structures and approaches to planning, for a wide variety of purposes.

In this corporation, and in this industry, Powerpoint is the usual medium for meeting fodder, although thankfully not for reports. I suspect that the distraction of unfamiliar approaches would raise a barrier to communication. Successful usage of Powerpoint seems to involve discussion ranging back and forth over just a few slides: a bulleted agenda or list of issues, and some diagrams. The place of the handout is taken by the e-mails and few-page memos that fulfil actions arising. Our corporate identity people have shown admirable restraint by giving us a plain white slide, discreetly but clearly numbered, with Ppphluff limited to one logo and some control info in small print across the bottom 10%. Of course, that does not prevent PP-excess: Marketing's current "reference customer presentation" in my subject contains almost 200 slides accumulated over 18 months. They expect sales teams to select as required for their pitch. Impossible!

Plain words, Tufte-guided graphics, and a Mosteller teaching mindset are starting to work for me. Feedback from a recent conference presentation was: "Many thanks for your time and energy to get me up to speed with (the subject). I've just been chatting with the other journalist I mentioned. Like me, he'd never got his head round this, thanks to cr*p explanations by pr/marcoms people, but is now very interested in learning more about it. I was also chatting about it with friends. They'd had a similar experience and once I explained how it worked and its benefits, they were also interested in hearing more...". What more could a communicator wish for?

The next conference will be a challenge: making a tricky technical subject interesting to an audience perhaps not so interested, right at the end of the last day. Definitely time to prepare a handout-style paper for the document pack, rather than just the slides.

-- Chris Horton (email)

Posting the differences between the first and second printings is a bit complicated because the main additions and significant changes are scattered over 9 different pages. Indeed there are minor changes on nearly every page of essay. Thus no "clean install". Here are the 5 pages that now discuss the Boeing PowerPoint slide and the section of the report of the Columbia Accident Investigation Board dealing with my analysis. That is followed by a question-and-answer postcript added in the new edition.

And then, placed at the end of the new essay, this postscript:

-- Edward Tufte

Analysis of the Columbia slide should be made only at

Analysis of PP should be placed in one of the many PP threads.

This thread is about the new edition of CSPP.

-- Edward Tufte

I had been giving technical and popular presentations to audiences for many years before Powerpoint came along. I was dismayed when I first tried to use it, but couldn't identify the source of my uneasiness until I read your analysis.

But I have developed a course of action for dealing with PP: I write my reports and papers in full prior to presentations, and then scan or otherwise import them into Powerpoint slides. In other words, I do use Powerpoint, but only as a computer-slide projector for regular report-format presentations. As for low graphics resolution, I tend to use overhead transparencies for material that requires high resolution. Sometimes I use transparencies for the entire presentation, as well.

-- Frank Sanders (email)

I swear this is the absolute truth: I heard a reporter on a New York City radio station today say that Yankee owner George Steinbrenner

"...collapsed while watching a PowerPoint presentation."

Unfortunate grammar? Or perhaps PowerPoint is more dangerous than we thought?

-- David White (email)

A recent review of the Cognitive Style of PowerPoint in Science (30 January 2004, volume 303, p. 630) by the English biologist Michael Stumpf.

-- Edward Tufte

I have just read "The Cognitive Style of PowerPoint" and found it enlightening and entertaining, though I do have a concern about the table on page 22, "Character Counts and Density Per Page-Image." The data set can be misleading because it includes unlike data environments (printed page and computer monitor display) with a like comparison (Characters/in^2). You pointed out in your November 9, 2003 response to this thread that the character counts were made from printouts from Netscape 7.0. This methodology, though expedient and unifying, does not, in my opinion, accurately quantify the comparative densities of a webpage v. printed page.

The main difficulty in comparing the two media is that one, the webpage, is a relative environment. Putting aside the display differences between browsers, browser versions and operating systems, and the printing differences encountered with different printers and print drivers; webpage density can be affected by the design decisions made by the developers. For example, uses absolute widths to display their tables, whereas uses relative widths. These design decisions make a difference in display density as well as printed density (and accuracy -- I printed off both of the above mentioned pages: the page filled 4 1/3 8.5"x11" pieces of paper, the page did not fill 2 pages, and the content to the extreme right of the page was excised because the absolute width was wider that the printable area of the paper).

To compare the relative digital medium, brought into the physical world by way of a desktop printer, to a physical medium, such as a page from the PDR, can be especially misleading if the single page of the relative medium (such as represents less than one quarter of the content of the page in its native environment. Understanding that the table's purpose was to show the paucity of data available on a PP slide, wouldn't a better measure of density be an expression of the character density in relation to a percent of available or usable area? While this would still be somewhat inaccurate when referencing webpages, it would provide a closer comparison of the disparate media. Or, in future revisions and editions, could you provide a description of the comparative environment (browser, OS, printer) for the data relating to the webpages.

-- k.michael.f.f (email)

The table mentioned immediately above is in the section of the essay called "Creeping PowerPoint: PP Slide Formats for Paper Reports and Computer Screens." This section deals with how the PP format has metastasized into a format for paper documents and internet screens. The resolution performance for best-selling books, news sites on the internet, and PP on paper and on computer screens is then compared by counting how much information each format depicts per page-image. In summary, the data indicate that the PP format used on paper runs 2% to 10% of the typographic richness of non-fiction bestsellers, and that popular news sites on the internet show 10 to 15 times more information than a large and diverse collection of PP-formatted images.

The details of standardizing computer screens are described above in this thread in the discussion by Jeffrey Berg and myself. I printed out the news sites the way I always print out screens from the internet, using the scaling and typographic conventions I've used for several years. If ridiculous tinker-toy formatted screens are used (following the Microsoft IE style!), the results will still yield approximately a five-fold reduction in information capacity per image from the use of the PP format. Thus the overall conclusion remains: "The PP slide format has probably the worst signal/nose ratio of any known method of communication on paper and computer screen. Extending PowerPoint to embrace paper and internet screens pollutes those display methods."

This analysis is consistent with the recent "lessons learned" by the new NASA safety and engineering group (quoted from GOVEXEC.COM )


By Beth Dickey

The reports from the NASA Engineering and Safety Center, based at Langley Research Center in Hampton, Va., are laced with reminders about managing risk, listening to dissenters, certifying flight hardware and limiting workloads. "We will be communicating the lessons and assigning actions across the agency," NESC director Ralph Roe told reporters Wednesday.

Roe briefed reporters after sharing the results of four studies with senior NASA leaders in Washington. The approach, actively sharing lessons learned, is borrowed from the U.S. Navy's Board of Inspection and Survey.

The initial assessments covered a problem with the space shuttle's rudder and speed brake system, the Mars rover landings earlier this year, the record-breaking flight of an experimental NASA aircraft in March, and a joint science mission with the French space agency.

On the French mission, called CALIPSO, the safety center looked at concerns about possible leaks of highly reactive spacecraft fuel. It held the flight of NASA's X-43A hypersonic craft until one team member's dissenting opinion about the vehicle's aerodynamic characteristics was addressed properly. On the Mars rovers, the safety center reviewed spacecraft instrumentation and plans for around-the-clock mission staffing.

Along with four technical reports, the NESC produced a four-page newsletter summarizing the technical activities and some lessons learned. The biggest lesson, Roe said, is to curb the practice of "PowerPoint engineering." The Columbia report chided NASA engineers for their reliance on bulleted presentations. In the four studies, the inspectors came to agree that PowerPoint slides are not a good tool for providing substantive documentation of results. "We think it's important to go back to the basics," Roe said. "We're making it a point with the agency that engineering organizations need to go back to writing engineering reports."

The NESC was created to serve as a source of expertise for evaluating the merits of technical concerns identified by agency employees. Its funding is not linked to any single NASA program or project, making it free from schedule or cost bias.

The center employs about 60 people full-time at NASA's Washington headquarters and 10 field installations in Alabama, California, Florida, Maryland, Mississippi, Ohio, Texas and Virginia. Up to 200 others from government agencies, industry and academia provide part-time support for individual cases.

Roe said his organization is one of the best examples of NASA's changing safety culture. Drawing experts from across the space agency breaks down the "stove-piping" compartmentalization tendencies criticized by the Columbia Accident Investigation Board, and the center's work is independent.

The center tries not to duplicate efforts by NASA program teams. But when parallel analyses are done and the data differ, Roe's NESC rules. "We have the authority to constrain a milestone or stop a flight or stop using a piece of hardware," he said.

In six months, the NASA Engineering and Safety Center has handled 49 requests. About one-fourth were generated by the space agency itself. The rest came independently from engineering and safety experts, senior managers, and at least two anonymous tipsters.

Although the center's $45 million budget has been "adequate for ramping up," Roe told reporters it may not be sufficient for NESC to handle future requests. They are being submitted at a rate of 200 per year, and "I think we will need a larger budget for this," he said.

The NESC is not to be confused with the independent technical engineering authority that Columbia accident investigators instructed NASA to establish. The agency's chiefs of engineering and safety still are working out the details for separate authorities at each NASA location. Roe's office will collaborate with them.

-- Edward Tufte

From the new edition, page 25: "PowerPoint allows speakers to pretend that they are giving a real talk, and audiences to pretend that they are listening." Chris Argyris, also of Yale (although now at Harvard Business), discusses this phenomenon generally: he talks about skilled incompetence (the skill of never upsetting anyone ever, whether it's helpful or not) and not talking about the undicussability of undiscussible issues, namely issues that may be threatening or embarrassing. So, he argues, you should expect any presenter in certain organizations to make powerpoint slides assuming that everyone wants to see them and assuming that, since powerpoint is what everyone wants, then they're all happy with the poor quality of output. At the presentation the prophesy becomes self-fulfilling because no one wants to threaten or embarass the presenter by asking for the real data, or demanding higher quality presentation materials. In fact, no one discusses the fact that no one would ever bring up such embarassing or threatening concerns.

So, in response to the FAQ style question in ET's postscript (page 26), "I work in a large bureaucracy and everyone uses PowerPoint. I have problems with PowerPoint but how can I avoid it in my talks?" I'd say, in addition to ET's advise, other references include Argyris's article "Skilled Incompetence" from the September-October 1986 issue of Harvard Business Review and any of Argyris's subsequent books. I know Organizational Learning II and Overcoming Organizational Defenses include discussions of this. Particularly in Overcoming Organization Defenses Argyris discusses a novel case study method to educate members of an organization about these organizational problems and personally develop methods to overcome them within that organization.

-- Niels Olson (email)

It seems to me that part of the problem is that Powerpoint is "just there," making it easier to use than to consider choices. Is anyone aware of any organization, such as NASA, that has gone to Microsoft and asked for either a seriously revised tool, or a version of Office that doesn't include Powerpoint?

-- Adam S. (email)

I thought ET and readers here might enjoy these clips about PowerPoint in government decision-making from "Will Iran be next?", The Atlantic, Dec. 2004. The article describes a war game exercise about the options a US president will have when if/when Iran goes nuclear.

"[Gardiner, the war games orchestrator] commitment to realism extended to presenting all his information in a series of PowerPoint slides, on which US military planners are so dependent that it is hard to imagine how Dwight Eisenhower pulled off D-Day without them. PowerPoint's imperfections as a deliberative tool are well known. Its formulaic outline structure can over emphasize some ideas or options and conceal others, and the amateurish graphic presentation of data often impedes understanding. But any simulation of a modern military exercise would be unconvincing without it....Gardiner's presentation used PowerPoint for its explanatory function and as a spine for discussion, its best use...."

Later on,

"When the exercise was over, I told David Kay that an observer who had not often seen such charts remarked on how "cool" they looked. "Yes, and the longer you've been around, the more you learn to be skeptical of the 'cool' factor in PowerPoint," Kay said. "I don't think the President had seen many charts like that before," he added, referring to President Bush as he reviewed war plans for Iraq."

The article concludes with some scathing remarks on how decision-making at the highest level these days seems to show streaks of a "PowerPoint cognitive style".

The consequence of PowerPoint might be more "world-changing" than engineering disasters like Columbia.

I enjoyed your class (I attended the Palo Alto one). Thanks!

-- deepak kenchammana (email)

My essay is mainly about PowerPoint, but it is also about the intellectual and moral failures of the pitch culture.

-- Edward Tufte

Science Fairs Start to Adopt PowerPoint: Modernization, Regression or a Lateral Move?

Prof. Tufte and the legion of thoughtful supporters in this forum (amongst whom I am a recent convert) still have some work to do as this New York Times article shows.

(Apologies if the link disappears behind a subscription fee requirement.)

The article discusses a move amongst some high school science fairs to use PowerPoint presentations instead of 3-fold cardboard displays. I'm struck by how neither method necessarily allows presenting a science experiment in an educational fashion, which surely must vary by type of information being reported by the experiment? I would have thought the ideal would a mix of methods based on the science being presented (models, displays, presentations, papers, etc.), rather than using only one method. Or does the three-panel display format introduce necessary structure?

[link updated February 2005]

-- Peter Morgan (email)


Like the scientific poster presentation of a research report, the cardboard display has become the standard format for science fairs. These events have been infected with the "cognitive style of PowerPoint" for years, and now they're finally going all the way. It's a little sad, really, since some kids can't even afford to buy the poster board (let alone a computer) and have to cut their display from used cardboard. The gap between the rich and poor grows...

-- Ted (email)

I have resisted contributing to this thread because I hate to strike a negative note, but if I'm honest I don't really agree with most of the objections raised against PowerPoint. Like everything that ET writes this article is stimulating and interesting, and I have learned some important things from it, but still, at the end I remain unconvinced about the main point.

First I want to get an irrelevance out of the way. Yes, I dislike PowerPoint for the same reasons that I dislike most Microsoft programs -- top-heavy, clunky to use, infuriating when it presumes to guess that I mean something different from what I've written, etc. For those reasons I gave up using it about a year ago and use Keynote instead. Others use Acrobat for similar reasons. However, all of this is irrelevant, because you can make bullet-laden low-intensity slides just as easily with Keynote or Acrobat as you can with PowerPoint.

Bullets. The problem that I have with this discussion is different. Maybe my experience is different from that of others in this forum, but I rarely see presentations full of bullets, rarely enough that although I don't like them any more than the rest of you do they don't particularly upset me when I do see them. Now I never attend military briefings, and rarely attend commercial presentations, and that way I doubtless miss a great many bullets. Most of the presentations I see are research seminars in biochemistry or molecular biology, but it is obvious from other contributions that there are others in this forum who work in similar fields to mine, and I wonder if they have the same experience of presentations.

Low density of information. So far as the printed page is concerned, I have long been totally in agreement with ET that a high proportion of graphics convey far less information than they could, that they use colour because they can and not because it necessarily adds anything useful, etc. There is thus no argument about that. The screen is another matter, however. Anyone old enough to remember pre-PowerPoint days must surely remember totally unreadable slides that tried to include about a printed page's worth of text on one slide. How many times have I heard a lecturer say "You won't be able to read this, but ..."? And such lecturers were not talking about what people at the back of the room could read with other people's heads in the way, but what people in the front row could read. Many times also one used to read advice along the lines of "Don't put too much on one slide". In an ideal world with perfectly focussed high-quality projection and everyone sitting in seats with an unobstructed view it might work to put a large amount of detail on one slide, but I've seen it go disastrously wrong so many times that I doubt it. Most people I talk to about this agree that slides are on average far easier to read now than they used to be before PowerPoint. It may well be that PowerPoint has taken things to the other extreme, but I think one has to recognize that there was a real problem in the past.

Linearity. One of the complaints made about PowerPoint is that it enforces a rigid predetermined sequence, and does not allow a dialogue between the speaker and the audience to decide the order in which the information is presented. Probably true, though it was no less true of 35mm slides. In principle overhead transparencies are better, but only if the speaker is sufficiently well organized to find the right transparency at the right moment. The blackboard, of course, is far better, and allows the discussion to take any direction at any moment. I nearly always prefer to use the blackboard for student classes for that reason, but I don't find it very practical for most research presentations. PowerPoint does allow links for jumping from one point to another, but I have never seen anyone (apart from myself) use them. I have rarely used them myself, mainly because people rarely make the sort of interruptions that would make a jump appropriate. (This may be a transatlantic thing: I have the impression that audiences in the US are far more likely to start a dialogue in the middle of a lecture than European ones are.)

Handouts. In principle it's a good idea for everyone to go away from a lecture with a paper handout provided by the lecturer. In practice, however, I rarely follow this advice, probably for bad reasons -- not wanting to deviate too much from what other people do, not wanting to carry great quantities of paper with me when I travel, etc. Probably the less said about this the better, because I can't really defend my practice on this issue. Nonetheless, I'd be interested to know how many of the other contributors to the forum put the recommendation into practice.

-- Athel Cornish-Bowden (email)

The conversation, above all, has been helpful: understanding information design allows me to analyze PowerPoint and other means of information delivery in meetings, briefings, and classes. ET's books occupy precious real estate next to Strunk and White and various style manuals and dictionaries in the shelf over my desk.

Bullets. I seem to get better grades in classes in which I spend more time on the class and in which I spend more of the time I allot to writing things down and drawing connections. Duh, right? Well, I've had professors who hand out the PowerPoint slides in advance. If I leave such a faux hand-out (FHO) turned up, I don't write as much and my grade generally isn't as good. I have, even before seeing this forum, put these FHOs face down to force myself to take notes. Similarly I don't seem to take as many notes when bullet lists are on screen. I think this has something to do with me thinking "Yes, what he's saying agrees with what's on the screen, and the whole thing seems like common sense. No need to write that down." In a class where the professor is writing on the chalkboard I write down, at a minimum, everything the professor writes, usually quite a bit more in a science class. Comparing notes across classes it appears professors and presenters don't use as many bullet lists when they don't use PowerPoint. The Columbia post-mortem's argument that bullets inhibit good logic by inhibiting good grammar and syntax seems sound, but I don't have observable results to add. There's also one massive risk to the chalkboard: if the student falls behind (I had one professor who would face the class while writing equations on the chalkboards around all four walls and talk about something other than what he was writing) and the chalkboard gets erased, then the student drops sync and can't recover the lost information even if the presenter slows down later.

Low density of information. Clearly the printed page presents more information in the same time, allows for more connections, and has more permanance. As a presenter and a student I prefer handouts. I think part of the information density argument goes to design and the data-to-ink ratio. ET also uses Keynote as a slide projector in his talks. I still think that the pixel count and other resolution factors are significant, but the data-to-ink ratio is where the biggest gains can be made. This goes for forms as well. I attend some lectures at a medical school in DC and they provide a feedback form to the audience members. Once I filled it with suggestions to increase the query-to-ink ratio of the form. I don't know who maintains the form, but that person made the changes and more people appear to turn the forms in.

Linearity. I disagree that transparencies are better for reducing linearity. I spent a year in a class in which the lecturer only used transparencies; it was just as linear as PowerPoint. That became particularly frustrating because the hand-drawn transparencies followed the text page for page, example for example. That was truly a case in favor of the students who don't think going to class is useful. The only advantage of the transparencies was resolution, which was helpful but was also balanced by the washed out colors of erasable markers. Personally, seeing the same material in two formats provided me an opportunity to compare the presentation styles of transparencies and textbooks, but it didn't help me learn Organic Chemistry.

Handouts. I use handouts; I don't use slideware. I don't have much feedback except an observable increase in dialogue and greater personal comfort while presenting. I will say just this afternoon someone came and asked me for the hand-out I made for a meeting last Wednesday. My office also owns a shredder the size of a small car and I cringe every time I make copies because it reminds me of driving through the Cascade mountains around Seattle about ten years ago: giant stumps as far as the eye could see.

-- Niels Olson (email)

A fascinating booklet, full of ideas that I will use in my teaching. Thank you.

Small suggestions for the next printing [page numbers below refer to the Nov 2004 edition]:

1. In the analysis of the Columbia slide (pp. 8, 9), you could help the reader by drawing an arrow from each of the "3 typographic orphans" (p. 9) to the location on the slide. Same for the ghastly volume measures (p. 9): arrows pointing to the locations would be helpful.

2. The arrows from the text blocks into the slide might be easier to track (and might seem visually distinct, to indicate their conceptual distinctness) if they were a different color, say red?

3. "Shakiness in conventions for units" (p. 9) should indeed provoke concern. An example is the Mars Climate Orbiter disaster in 1999 (see <> for a detailed discussion of what went wrong).

4. Why not write dates as 18 March 2003 (footnote 7 on p. 11) instead of March 18, 2003? The first method uses less punctuation, making bibliographic information easier to read.

5. The name PowerPoint is, as you say, about making "power followers" (p. 13). A related point, so to speak, is that it creates a single point of power, the speaker, which destroys the possibility of dialogue -- which requires multiple poles -- and hence of teaching.

6. Shouldn't the footnote tags that begin the actual footnotes be in the same font size as the footnote text, and not superscripted? In the text, the footnote number should be small and superscripted so that it doesn't interrupt the flow too much, but in the footnote text that reason does not apply, and I find the superscripted number hard to spot.

7. Use single quotes instead of double quotes, even though American publishing style prefers double quotes. The double quotes lead to moth-eaten margins: e.g. the last footnote on p. 16 in the line that begins ``Replacement of the Aortic Root; or the line (also p. 16) in the last text paragraph that begins ``Sweet songs never last too long. The white space below the `` looks like it was eaten by a mouse or a moth. And with a single quote, you could even use hanging indentation, putting it in the margin, which would protrude too much with a double quote. Single quote also allow commas and periods to fall inside or outside according to meaning (inside clauses but outside titles such as `Instructional Computig Facility' [p. 19, 2nd paragraph]), whereas double quotes force them inside always -- a loss of fidelity!

8. The 2nd to last sentence on p. 24 ends strangely: "that you seek to leave traces and have consequences." The 'have' rings funny. Did you mean 'accept'?

9. Do tables in _Science_ really have 1000+ entries in them? That number sounds a bit large. I don't have a copy easily to hand, but I will look when I am next near one.

10. Footnote 15 on p. 20 (saying that n cells yield n(n-1)/2 pairwide comparisons) is accurate, but a less accurate and more useful formula is n^2/2. That form shows easily that splitting a table of n elements into k subtables (e.g. Gaunt's n=1,855 counts into k=155 slides) gives k*[(n/k)^2/2] comparisons, which is (n^2/2)/k. So the 155 PP slides give, in total, 1/155 of the number of comparisons, or roughly 6 in 1000 (as you say). But with this approximate method, you do not have to find the huge number of pairwise comparisons in the two alternative methods and divide them: You get the comparison (which is all that one may need) right away.

11. One point you make, that we should provide high-resolution handouts and _talk_ through them with the audience, reminds me of two excellent papers on the "Gutenberg method" of teaching (i.e. teaching that takes account of the invention of the printing press):

- Morrison, R. 1986. "The Lecture System in Teaching Science," in "Undergraduate Education in Chemistry and Physics: Proceedings of the Chicago Conferences on Liberal Education," No. 1, edited by R.R. Rice (Univ. of Chicago), p. 50-58. [Excerpts from Morrison's wonderful article are at <>]

- Lambert, F.L. 1963. "Editorially Speaking: Effective Teaching of Organic Chemistry," J. Chem. Ed. 40: 173-174.

[Thanks to Richard Hake for introducing me to these two gems.]


-- Sanjoy Mahajan (email)

Thanks so much for these helpful comments.

The median size of the data matrix lurking behind a graphic in Science is more than 1000 numbers; the analysis is not about published tables but rather the data matrices behind published graphics.

-- Edward Tufte

Have you considered making the text version of the above pages from the booklet available in the forum, so as to enable indexing by both external search engines and your internal search engine?

-- John Driscoll (email)

Of course, the universe of discourse for this discussion assumes that the person making a presentation wishes to inform the audience of something (teach, educate, enlighten, communicate). And that the audience are expecting to learn something from it.

But in reality the vast majority of corporate communications are sales communications. Of course, the sales guys try to sell to potential customers. But at the same time, the purpose of a project update is not actually to update the higher-ups on the status of the project, present problems for their guidance, and ask for decisions where they are needed. The real purpose of these meetings and presentation is to sell to the higher levels of management the perception that the person/department making the presentation is doing a fabulous job. Perhaps the actual information is negative; then the message is "despite the bad news, here's what I am doing to turn the situation around." Or if the news is basically good: "Look what my efforts have achieved." And of course, since most complicated business information can be interpreted in a variety of ways, "What this confusing information really means can be summarized best by me in a way (without too many confusing sentences, or actual graphs, but with lots of arrows and animations) that shows you that things are really going along well, thanks to me and my team." The presenter who can make the most convincing sales pitches to higher levels of management is the one who will advance farther and faster than the others.

So, it is beside the point that PP is terrible at communicating. What its users and their audiences are using it for, is sales, pure and simple.

-- J. Calhoun

I am a student at the University of North Texas and attended your seminar in Austin TX last month. I very much enjoyed your Cognitive Style of PowerPoint article. It really articulated what I have hated about PowerPoint for so long but was unable to say myself. That it is a crutch, data thin, and should be boxed into a wooden crate and fired into the heart of the sun.

I thought you might want to see a horrible PP presentation from the Dr Mikler in the UNT College of Engineering on Probability. It is complete with mountain/moon scape and Comic Sans font. Fortunately I had taken a proper Probability class from the math department, unlike others in the class.

Thank you so much for your books. I finished Visual Display a few weeks ago.

-- Cameron Palmer (email)

The "PowerPoint Cognitive Style" is not limited to (Microsoft's) software. Digital projectors, and especially their feature-laden controllers, are emulating the PP styles with similar embedded hierarchies and feature clutter. I've witnessed many failed presentations as speakers sadly fumble with the maze of controller buttons in a darkened room. The wrong button is pushed and images spin out of control, recurring default screens, and, finally, the "can't read file" alert. This is especially troublesome for those speakers (myself included) that are simply trying to use the digital projector as an image display. The abandonment of slide projectors is nearly complete leaving few options (yes, handouts and clear public speaking is preferred, but sometimes images are a necessary adjunct). There is certainly a market for more simple digital projectors capable of displaying images analogous to the function of a slide projector. Do any "simple" digital projectors/controllers exist? Thanks,

-- Daniel Meatte (email)

Kensington makes a very simple USB cordless controller with laser pointer and four buttons. It is Mac and PC compatible and ~$40 if you shop around. I use one with in lecturing to 400+ biochemistry students and in seminars. Similar products are available from Logitech and other manufacturers.

-- Alexey Merz (email)

Athel Cornish-Bowden writes above, responding to the criticism of Powerpoint as "low density of information", that many older presentations packed an unreadable quantity of text onto a single slide.

I would say that information is not text, so a slide's data content could be much improved without necessarily including much more text, and that text is not information, so those illegible old text-dense slides were information *poor*, not information rich.

I remember pageful-of-dense-text slides (well, I still see them today), and they struck me as having had a function for the presenter similar to the function of many more trendy slides: not to give the knowledge the presenter has to the audience, but to assure the audience that the presenter has the knowledge he claims to have :-)

-- Derek Cotter (email)

Just published is the 2nd edition of The Cognitive Style of PowerPoint. It is now 32 pages long; the original essay was 24 pages. The new edition contains the material on PowerPoint and rocket science that opens this thread, a long discussion of the causes of presentations (sorting out variation among users, content, and presentation methods), and an essay on lists. This new edition is also a chapter in Beautiful Evidence.

For more information and to order click here.

-- Edward Tufte

I just received my copy of the new edition. On page 14 you say that "Replacing PowerPoint with Microsoft Word (or, better, a tool with non-proprietary universal formats) will make presentations and their audiences smarter."

The XML-based OASIS Open Document Format for Office Applications (on its way to becoming ISO/IEC 26300) may be that format, and is a tool that supports that format. [See for details about OpenDocument.]

-- J. David Eisenberg (email)

German conference on Powerpoint

A whole conference on Powerpoint. Unfortunately I don't read much German.

From: Claus Pias <> Date: 1. Juli 2006 Subject: CONF: Powerpoint. Prasentieren in Wissenschaft und Wirtschaft

Powerpoint. Prasentieren in Wissenschaft und Wirtschaft

5.7.2006, 10-20 Uhr, Humboldt-Universitat zu Berlin Hauptgebaude, Unter den Linden 6, Demoraum des CMS, Erdgescho??

Wissenschaftliche, technische und kommerzielle Kommunikation nutzen zunehmend visuelle Elemente. Zur dominierenden Prasentationstechnik hat sich Powerpoint entwickelt. Dies fuhrt zu einer eigenen "Prasentationskultur", deren Erscheinungsformen, Eigenheiten und Widerspruche in einem Workshop beleuchtet werden sollen. Ist Powerpoint nur eine Durchgangsphase vom vorgelesenen Text zum projizierten Bild oder verandern Multimediatechniken die Art und Weise der wissenschaftlichen und technischen Prasentation nachhaltig?


ab 10:00 Einla?? und Anmeldung

11:00-11:15 Begru??ung

11:15-12:00 Prof. Claus Pias, Universitat Wien, Image Supported Cooperative Work. Zur Vorgeschichte von PowerPoint

12:00-12:45 Prof. Klaus Rebensburg, Technische Universitat Berlin und HFF Potsdam, Powerpoint: Von Kraftpunkten uber Kraftlosigkeiten zu Katastrophen

12:45-14:00 Mittagspause

14:00-14:45 Prof. Jorg-Martin Pfluger, Auf den Punkt gebracht. Demonstration und Narration

14:45-15:30 Prof. Stefan Romer, Akademie der Bildenden Kunste Munchen, Diagrammatische Rhetoriken - Powerpoint als Zwangspadagogik

15:30-16:00 Kaffeepause

16:00-16:45 Dr. Claus Noppeney, CNC Berlin, >Ich wurde Folien von mir nie als Kunst bezeichnen, obwohl ich sehr schone Folien mache.< Beobachtungen zu PowerPoint in Unternehmen.

16:45-17:30 Prof. Wolfgang Coy, Humboldt-Universitat zu Berlin, Vortragstechnik.

17:30-20:00 Abschlu??veranstaltung und geselliges Zusammensein

Um formlose Anmeldung wird gebeten unter: / Fax 030-484985013

Ein Workshop des Berliner Stiftungsverbundkollegs der Alcatel SEL Stiftung Organisation: Wolfgang Coy / Claus Pias

-- Matthew Gilmore (email)

On the lighter side, view this YouTube video titled "How NOT To Use Powerpoint By Comedian Don McMillan". This has been circulating around our office since our corporate culture relies heavily on Powerpoint.

-- TShaffer (email)

Clearly ET has struck a chord with his original, and now the second, edition of "The Cognitive Style of PowerPoint". A manner of presentation that is now routinely taught to elementary schoolchildren in districts wealthy enough to afford students computer access should give adults pause before use anyway. PowerPoint is to traditional, data-rich tables and charts as the newspaper USA Today is to traditional newspapers. My comment is no accident, as the comic strip "Doonesbury" lampooned years ago the tendency of USA Today to use bullet points and brief lists. PowerPoint is a cheesy, dumbed-down method of attempting to convey data and information, even for this postliterate era. I can't count the number of times I've seen misspelled words, sloppy phrasing, excessive use of colors and typefaces, and inappropriate layout in slides created with this program. Those of us who are old-school owe ET a debt of gratitude for this work.

-- Ed Krampitz, Jr. (email)

Today's online edition of The Sun (UK) features some phluffish graphics of the delegate counts of the US presidential candidates:

-- Michael Cusack (email)

Threads relevant to PowerPoint:
Don't get your hopes up.
Compares tables, slopegraphs, barcharts for showing cancer survival rates.
A look at a rich and complex question: What are the the causes of presentations?

An intriguing but under-explored topic.
Account of the role of PP in the shuttle Columbia accident, followed by many good alternative methods and examples for technical presentations.
Mainly recent examples of leaked PP slides in the Iraq war.