The Cognitive Style of PowerPoint

Pitching Out Corrupts Within

by Edward R. Tufte

2nd ed., 32 pages, Graphics Press LLC, Cheshire, CN, January 2006.

Reviewed by J. M. Haile, Macatea Productions,

The evolution of storytelling stretches from prehistory to the present day, for humans love stories. We tell and listen to stories every day, and though the settings and motivations differ, we judge a story to be effective when it meets some standard for belief. Those standards differ depending on the kind of story being told.

Cultural stories may be told to entertain, to inspire, to instruct, to preserve a culture's history, to define cultural norms. A rich cultural story will intertwine several of these objectives simultaneously. Cultural storytelling is invariably fiction, but through fiction we seek to illuminate some kernel of truth. Even when based on "true" events, modern film remains fiction: this is necessary to satisfy the time and economic constraints of film-making and to satisfy the dramatic requirements that propel a story forward. Part of the art in film-making is to fictionalize in such a way that the underlying truth remains undisturbed. But in any case, even when a film is complete fantasy, we require "believability." Much of that believability rests on the internal consistency of events taking place within the film's limited universe, in particular, on whether the behavior of the characters is consistent with the events that motivated them.

Stories of the marketplace are told to persuade: the presenter intends to bring the audience to his point of view. Here, believability is measured by the credibility of the presenter and success is measured by the extent to which the audience is convinced.

If evidence is available to support the presenter's position, it will be used. If no evidence is available, the presenter will appeal to "apparently reasonable" arguments and to the psychological needs of the audience. If the available evidence is against the viewpoint, the presenter may either ignore the data or belittle it and appeal to the audience's self-interest.

But technical stories—those told in science and engineering—invoke believability of a different kind: the arguments for belief must be rooted in observations—preferably experiments in the physical world—and the chains of reasoning from those observations must obey the laws of logic. Usually, such reasoning is realized in the language of mathematics, for mathematical reasoning is at once economical, precise, and universally accepted. The goal of the presenter is to evoke belief using evidence and reasoning that is ultimately unquestionable, so that the story must be accepted, even if the audience is predisposed otherwise. It is crucial to note that technical presentations make no appeal to persuasion—no appeal to the psychological states or self-interests of the audience—and they avoid fiction like the plague.

Over the last hundred years or so, we have seen a happy marriage of cultural storytelling with technology: film, cameras, sound recordings, projection equipment, sound reproduction, and now computers and digital processing have combined to make forceful impressions on human senses. Such a combination of technology and human imagination may not merely entertain, but also enlighten; conceivably, it might even expand human awareness.

To a lesser extent, we have also seen a union of marketplace storytelling with technology—the pervasive form of that technology is now Microsoft's PowerPoint software. Whether this union is, in fact, a happy one probably depends on who is asked. Certainly, it is happy for Microsoft and for presenters, generally; but it is likely viewed as a mixed bag by many audiences.

But here is an irony: while technological developments have drastically amplified the impact of cultural storytelling, and they have certainly enabled new ways for persuasion in the marketplace, technological developments have been little exploited to improve technical storytelling. And so, it is perhaps understandable that some scientists and engineers, in efforts to appear up-to-date in the technology that they in fact created, have been seduced to using the persuader's software as the vehicle for telling technical stories.

Edward Tufte's position is that this marriage is detrimental to technical discourse, and the sooner we can effect a divorce, the better. His arguments are devastating. Why might PowerPoint be the wrong tool for telling technical stories? Here are some of Tufte's points.

  1. PowerPoint slides are distinctly low resolution compared to other media. Most PowerPoint slides are text-only, and when graphics are used, the graphics often have nothing to do with content. Nevertheless, graphics occupy space on a slide, so to make fair comparisons, we consider here text-only slides. We speak at 100-150 words per minute and read at up to several hundred words per minute. The density of content in reference works (such as the Physician's Desk Reference, the World Almanac, the Guiness Book of Records) is between 150 and 200 characters per square inch; in news publications (such as the New York Times, the LA Times, and Google News) it is between 40 and 50 characters per square inch. But the average text-only PowerPoint slide contains less than 5 characters per square inch; that's an average of about 40 words per slide. Such meagerness is comparable to the "Dick and Jane" stories of preschool readers.
  2. PowerPoint forces a structural style on the content. In the case of PowerPoint's standard templates, that style is a hierarchical nest of lists—invariably with bullets—the power points. Such a structure may be suitable for those trying to persuade, but one-dimensional lists are not effective for making comparisons, evaluating evidence, or developing chains of logic. When we want to compare things, we prefer to see them side-by-side: visual comparisons are best done by separating objects in space. But the limitations of PowerPoint force comparisons to be done by separating images in time: the presenter must flip back-and-forth among images. This may be relatively easy for the presenter to do, but it is counterproductive for the audience. More generally, modern technical issues rarely resolve into simple hierarchical structures: modern technical problems are not simply one- or two-dimensional, numerous variables may be involved and important, quantities may be interrelated in nonlinear ways, and connections between causes and effects may be numerous and finely balanced. Forcing such complexity to fit a linear hierarchical structure can easily leave an audience with incomplete and erroneous understandings. Such misapprehensions may, in fact, be the goals of the persuader, but they are not the goals of scientists and engineers. The structure of the content should dictate the structure of the presentation, not the other way around.
  3. PowerPoint's bells and whistles are a mockery to serious thought. Tufte posits a vicious circle: (a) After preparing the slides, the presenter realizes the content is thin and the talk will be boring, so (b) he adds a few bells and whistles to add "interest", but (c) now the talk will be even more boring because the irrelevancies detract from the content, so the presenter is back to step (a) and after a couple of iterations, a potentially weak presentation becomes a disaster.

These are not by any means all of Tufte's concerns, but they are enough to give you the flavor of the argument. But in addition, Tufte goes on to explore how PowerPoint presentations are being used in technical settings. His examples come from NASA, but the practice is surely more widespread: in some technical communities, PowerPoint presentations are now being used as the sole bases for making technical decisions—replacing technical reports, white papers, and analytic documentations of studies and reviews of evidence. In fact, Tufte quotes a NASA report,

"It appears that many young engineers do not understand the need for, or know how to prepare, formal engineering documents … ."

This is an alarming statement, as well as a serious indictment of management in technical organizations and of teachers in educational institutions. An oral briefing on a technical issue stands to the underlying documentation in much the same way that a trailer stands to a motion picture, and decisions about the quality, content, and impact of films will be routinely erroneous if they are based solely on trailers.

What Tufte does not touch upon is the growing use of PowerPoint as a primary medium for educating engineers in universities. Again, the irony is palpable, for technical educators view with disdain and condensation the practices and ethics of the marketplace. However, Tufte does observe that good oral presentations employ the same cognitive style as good teaching. And good teaching does not reduce technical material to a simplistic Dick-and-Jane level. Good teaching needs no PowerPoint bells and whistles to capture student interests. Good teaching reveals the beauty, symmetry of form, and surprises in technical material that attract and stimulate human minds. A good presentation style, as part of a good teaching style, produces scientists and engineers who can fix technical problems, rather than ones who can merely fix the pitch about the problem.

If you give oral presentations in technical settings, if you teach students how to prepare and present oral reports, if you teach students science or engineering, you must read and ponder what Tufte has to say.

(jmh 03 June 06) © 2006 by J. M. Haile. All rights reserved.