Using Visual Rhetoric to Avoid PowerPoint Pitfalls

Transcription

1 Using Visual Rhetoric to Avoid PowerPoint Pitfalls Alan Manning Brigham Young University Nicole Amare University of South Alabama Abstract Criticisms that Tufte and others have leveled against PowerPoint are not insurmountable defects of the programs themselves. These defects are generally due to an orientation, shared by program designers and users alike, and toward images rather than diagrams, toward perceptual decoration and object indication rather than toward visually mediated, iconic representations of verbal information. Using Peirce s theories of visual rhetoric, we show that improvements in visual communication generally and PowerPoint slides in particular depend on shifting our orientation away from image-driven thinking and toward diagrammatic modes of presentation. Keywords: visual rhetoric, PowerPoint, graphics, Edward Tufte, Charles Peirce Critics such as Tufte, Parker, Norvig, and Jaffe [1] [2] [3] [4] argue that PowerPoint (or any equivalent slideshow program) causes those who use it to produce inferior visual presentations. Amare [5] counters that PowerPoint is a useful tool with some inherent weaknesses that users need to be keenly aware of: According to Microsoft, PowerPoint was developed to save the inferior speaker from embarrassing presentations and audiences from bad speeches. But Cliff Nass, a Stanford University professor, says that PowerPoint not only lifts the floor but it also lowers the ceiling. [2, cited in 5, p. 63] In other words, PowerPoint may indeed make the worst presentations better, but its more automatic and easily accessed features may tempt even skilled presenters to be less effective than they might be. We propose here to extend Amare s [5], Dumont s [6], and Keller and Shwom s [7] moderate critique of PowerPoint by developing a model of visual rhetoric that can be used to explain the tool s weaknesses and help thoughtful presenters to side-step and ultimately overcome them. The philosopher C.S. Peirce wrote back in 1904 that rhetoric should be understood as the art of making any kind of communication effective. Evidently, our conception of rhetoric has got to be generalized; and while we are about it, why not remove the restriction of rhetoric to speech? What is the principle virtue ascribed to algebraic notation, if it be not the rhetorical virtue of perspicuity? Has not many a picture, many a sculpture the very same fault which in a poem we analyze as being 'too rhetorical.' Let us cut short such objections by acknowledging at once, as an ens en posse, a universal art of rhetoric, which shall be the general secret of rendering signs effective, including under the term 'sign' every picture, diagram, natural cry, pointing finger...numeral, word, sentence, chapter, book... [8, p ] Textual rhetoric is thus the art of making effective texts, and visual rhetoric is the art of making effective visuals here we particularly mean visuals which support textual information tables, graphs, diagrams, charts, photographs, and illustrations of course, but also page layout choices, margins and borders, white space and /05/$ IEEE. 281

2 fonts. We can understand effectiveness only in terms of goals: what the author means to accomplish with this or that visual effect and what effect an audience needs when presented with a text supported with visuals. We can begin, at least, to understand author/audience goals by again following Peirce and sorting these goals into three major heuristic categories: (Figure 1) that images (such as photographs) or flashing pointers (such as LOOK OUT ) mediate between raw perception of form and the actual objects being indicated. (Figure 2) Figure 2. The mediating function of visuals. Images and pointers mediate between perception and action; informative visuals mediate between perception/action and language. Figure 1. Visuals sorted into Peirce s heuristic categories. Decorative visuals evoke feelings; indicative visuals provoke action; informative visuals promote understanding. Some types of visuals are highly effective for accomplishing one kind of goal but not others. We propose to use this three-part division of rhetorical goals as a means of evaluating and improving visual communication and PowerPoint presentations in particular: decoratives, indicatives, and informatives. The decorative aesthetics of a text have definite value, but novice authors tend to overdecorate with eye-catching fonts and photographs, for instance. PowerPoint presentations frequently overdecorate but also often overuse action-provoking indicative strategies. This is partly because the PowerPoint program itself is constructed to automatically bullet text and then to easily animate those bullets. The audience is moved to indicative actions, to follow bits of text as they buzz around the screen; but without clear goals, these actions become mere irritations. The key to improvement is an understanding of informative visuals: how diagrams and tables, for instance, differ rhetorically from photographs and moving bullets. In Peirce s framework, diagrams, charts, graphs, and tables represent iconic forms of communication that mediate between purely informative language and the raw perception of feelings and the raw interaction with physical things [9, paragraph 278]. This intermediate status of informative visuals is comparable to the way Consider, for example, our reaction to any raw perception, say, a light tickling feeling on the back of one s neck. Is this a good thing? Bad? Neutral? Whatever image instantly comes to mind mediates how we will react: the light touch of a lover or a loose thread from your collar or, perhaps, the picture of a free-roaming spider. Regardless of the real cause, the mental image determines whether we will sigh with pleasure or jump in terror. By the same token, according to Peirce, mental icons are either non-logical (=diagrams) or logical (=graphs or table-like comparisons) and always mediate both (1) our perception of and (2) our reaction to verbal information (Figure 2). In all primitive writing, such as the Egyptian hieroglyphics, there are icons of a non-logical kind, the ideographs. In the earliest form of speech, there probably was a large element of mimicry. But in all languages known, such representations have been replaced by conventional auditory signs. These, however, are such that they can only be explained by icons. But in the syntax of every language there are logical icons of the kind that are aided by conventional rules... [9, paragraph 280] In other words, the underlying meaning of any word, sentence, or any longer text must be understood as a diagram, either visually or in some other sensory mode, either consciously or unconsciously. This, in theory, is the fundamental 282

3 reason that we use visuals, in PowerPoint or anywhere else, when we present complex information. The meaning of any verbal information, if it is understood, has to be recreated in the minds of the audience as some form of iconic diagram or graph. Either the speaker can foist upon the audience all of this labor of reconstructing verbal information as visual information, or the speaker can at least partially process the information visually, in advance, on behalf of the audience. Realistic images may also help an audience to process information, but they do not have the same informative structure as diagrams, graphs, and tables. Peirce here is careful to distinguish between language-mediating icons (diagrams and graphs/tables) and raw perceptual/indicative images (such as photographs). Photographs, especially instantaneous photographs, are very instructive, because we know that they are in certain respects exactly like the objects they represent. But this resemblance is due to the photographs having been produced under such circumstances that they were physically forced to correspond point by point to nature. In that aspect, then, they belong to the second class of signs, those by physical connection. [9, paragraph 281] This image vs. diagram/graph distinction serves here as our primary focus, the core basis of our analysis/critique of visual communication in general and PowerPoint slides in particular. In the digital age, photographs are simple to acquire and paste into a visual presentation. If the goal of using the visual is decorative, to make the thing look nice, to communicate a specific feeling, or if the goal is indicative, to identify one specific individual, then we might expect that images will be preferable to diagrams. (Figure 3) On the other hand, images have a number of cumbersome properties if the primary goal is informative communication and if the visuals were chosen to support textual information. First, natural images have a general lack of clear contrasts: Consider for example the image of the spider in Figure 3, left, as opposed to the diagram, top center. Image details of the spider s body mix rather seamlessly with shadows on the rock face. Figure 3. Images vs. Diagrams. Images communicate feeling and indicate specific objects; diagrams inform by emphasizing contrasts, restricting detail, and generalizing concepts. The image, to be grasped, has to be perceived holistically. And yet information, to be understood, has to be clearly divided into contrastive parts: problem and solution, thesis and evidence, data and interpretation, argument and counterargument, etc. Second, as Peirce notes, images correspond point to point with things in nature, and as such, images have no filter for irrelevant details. Whatever contrastive details are relevant to an informative purpose are easily lost in the crowd. Again, compare the spider image above with either the diagram or graph, where only the critical features (main body-part divisions or body length vs. silk thickness) are represented. Third, by their nature, images focus on traits of an individual case rather than the general properties of a group. It is difficult to draw reliable conclusions from the image of just one particular individual if the goal is to communicate the general properties of the type of thing the image is meant to represent. International audiences are particularly likely to falter when presented with culturespecific images which to them do not serve as good prototypes for general concepts. For example, an image of a U.S. mailbox is commonly used to indicate functions, etc., and yet this image bears little resemblance to mailboxes in most other nations [10]. We can now show that the criticisms that Tufte and others have leveled against PowerPoint (or other 283

4 slide-presentation programs) are not insuperable defects of the programs themselves. Rather, these defects are generally due to an orientation, shared by program designers and users alike, toward images rather than diagrams, toward perceptual decoration and object indication rather than toward visually mediated, iconic representations of verbal information. Improvements in visual communication generally and PowerPoint slides in particular depend on shifting our orientation away from image-driven thinking and toward diagrammatic modes of presentation: Images: 1. lack of clear contrasts 2. no filter for irrelevant detail 3. unreliable generalization Diagrams: 1. clear contrasts 2. relevant details only 3. generalization reliable It is critical at this point to realize that NOT everything that looks like a diagram or table will actually function properly as an informative visual. Not all verbal text will function as information either, if it is presented in a form that an audience cannot process. This is the most certainly the case in Figure 4, a replica of Peirce s original table of ten basic sign types. [9, paragraph 264]. This table/diagram is the basis of Figures 1-3 in our own presentation, but our source figure cannot function informatively for most audiences. Without pages of discussion, without context, it s not possible to interpret Figure 4. The terminology is opaque ( rhematic, iconic, dicent and so forth). The significance of the arrangement of the boxes is unfamiliar. This lack of contextualization is intentional on our part. It shows what happens when an apparently informative graphic actually isn t informative at all. Precisely because it is not interpretable in its original form, Peirce s graphic is, for most audiences, only an image, a decorative-indicative replica of what Peirce s original chart looked like. So, even though it looks like it might be informative for someone, Figure 4 has all the expected properties of an image, for most audiences: (1) lack of clear contrasts, (2) no filter for irrelevant detail, and (3) the audience can extract no generalizations from it. To make our own Figures 1-3 workable, we had to reprocess the information in Figure 4. The Roman Figure 4. Peirce s original representation of sign types. Without adequate context, Peirce s table/diagram can only be processed as an image. numerals I-X in Figure 4 form an inverted pyramid, which we made solid in our own figures. Figure 5 indicates how we translated Peirce s original graphic. Figure 5. Reprocessing Peirce s representation of sign types. Peirce s schema is made intelligible by emphasizing key contrasts, filtering detail, and emphasizing a key generalization. In other words, we had to simplify and reconstruct Peirce s original visual so that it would function properly for a modern audience as diagram: (1) making the most fundamental contrasts apparent, (2) filtering out all but the most relevant details and (3) making apparent the general distinction between sign type II (images) and types V and VI (diagrams and graphs/tables). 284

5 PowerPoint problems can typically be traced to forms like Figure 4, forms that superficially look like informative visuals but which are actually images in terms of their effect on audience. Tufte cites examples of chartjunk [1, pp ] that clearly fit this description (Figure 6). Figure 7. Keller and Shwom s redemption of PowerPoint graphics. PowerPoint graphics tools can recreate content in diagrammatic form. Figure 6. Tufte s chartjunk. PowerPoint graphics tools make it relatively easy to create overcomplicated graphics that cannot be processed as informative visuals. In such overly complicated graphics, the audience is overwhelmed by the sheer number of distinct forms, dimensions, and data points. Threedimensional graphics are particularly problematic precisely because the eye wants to interpret them as actual objects, a wizard-hat collection or a traffic-cone convention (i.e. images) rather than as schematic relationships (i.e. diagrams). Though they resemble genuine diagrams, such forms have, once again, all the expected properties of an image: (1) Lack of clear contrasts there are so many contrasts that they all collectively lose their effect. (2) No filter for irrelevant detail there is no way to tell which data trends are significant and which are not. (3) The audience can extract no generalizations the overall point of the data is unclear. It is perhaps too easy to produce such graphics in PowerPoint, just as it is too easy to hit one s thumb when driving nails with a hammer. This does not mean that tool users cannot be trained to avoid such pitfalls. Keller and Shwom show that the same data from Tufte s chartjunk example can be reprocessed using PowerPoint s graphics generator, to create Figure 7, precisely the kind of graphic that Tufte recommends [1, p. 16; 7, p. 12]. As with the original Peirce visual, information density in Figure 6 is sacrificed for the sake of information clarity in the revision, Figure 7, which now has the properties of a diagram: (1) Fewer types of cancer are compared, and consequently the contrasts become more meaningful. (2) The graphic is made two-dimensional; only the relevant trends are shown. (3) General conclusions are apparent: e.g., survival rates for myeloma are good in the first five years, but relatively poor over 20 years, compared to other cancers, etc. A similar analysis can be applied to the use, or overuse, of bullet points in slide presentations. A bulleted list is fundamentally an indicative strategy, meaning the audience is provoked to actively focus attention on a slide feature, in this case to compare and contrast indicated items in a list. Consequently, the limitations of bullet lists are essentially the same as other indicative-contrastive strategies, such as the use of contrasting colors, or attention-grabbing fonts, or directional arrows: overuse of any such feature on any given slide destroys its effectiveness. As the contrasts blur together, the visual begins to take on image properties (like Figures 4 and 6), as opposed to diagram properties (like Figures 5 and 7). The conventional wisdom has been that long bullet lists are manageable if broken into hierarchies, say five main bullets with three or four sub-bullets each (making bullets in all). Our analysis suggests that even this is far too many. Seven not- 285

6 too-wordy bullets is probably the effective limit, and even then, these should be broken into two hierarchies of three and four lines, at most [11]. The great danger here is that the slide becomes filled with text; the audience usually has neither the time nor the inclination to read a full screen [6, p. 1-2]. As before, when any visual, even primarily verbal text, is not processed by the audience, then it does not really count as an informative graphic. Once again we have created an image of information, a decorative indication of information with unprocessed content: (1) contrasts blurred, (2) unfiltered, even overwhelming detail, and (3) the typical lack of general conclusions. (Figure 8, from [7, p. 9]) Figure 9. Bullet points as markers of hierarchical structure only. Bullet points are most effective as contrast markers if used sparingly. bullet points or graphics generators. Tufte and others are, therefore, incorrect in their claim that the cognitive style of PowerPoint somehow makes us stupid [3] and that slideware programs are inferior to handouts. PowerPoint users need to be aware of the common tendency to use visuals as decoratives and indicatives when informative visuals are more appropriate. We have proposed here an application of Peirce s visual rhetoric to explain the actual source of problems with graphics in slideware programs. We hope that technical communicators, keeping the image/diagram distinction in mind, will be able to take their PowerPoint presentations to a new level. Image Credits Figure 8. Hierarchical subordination of multiple bullet points. Bullet points, as indicative visuals, lose effectiveness if overused. In contrast, Farkas recommends that bullet lists only be used to mark out the overall hierarchical structure of a given piece of information. Figure 9 constitutes a diagram of the structure of the overall points in his article [12, p. 29]. As before, Figure 9 has the expected diagram properties: relatively few but clear contrasts, relevant detail only, and a clear general conclusion. We conclude that Tufte and others are correct in pointing out that PowerPoint graphics are often inferior, but the problem of presentation quality cannot be blamed solely on PowerPoint s low resolution or small screen size, nor on automatic Wolf spider photo: ungry_spiders.html Accessed Dec. 8, Spider parts diagram: ids/label/extanatomy/answers.shtml Accessed Dec. 8, Spider silk graph: lk_stealers.html Accessed Dec. 8, 2004 References [1] E. Tufte, The Cognitive Style of PowerPoint, Chesire, CT: Graphics Press,

7 [2] I. Parker, Ian. Absolute PowerPoint: Can a software package edit our thoughts? The New Yorker. pp , May 28, Available: ABSOLUTE%20POWERPOINT.doc [3] P. Norvig. The Gettysburg PowerPoint presentation. 18 Jan Available: [4] G. Jaffe. What's your point, lieutenant? Please, just cut to the pie charts. Wall Street Journal, 26 Apr. 2000: 1. [5] N. Amare, Technology for technology s sake: The proliferation of PowerPoint, in IPCC 2004 Proceedings, Minneapolis, MN: IEEE, pp [6] J. Doumont, The cognitive style of PowerPoint: Slides are not all evil, Technical Communication (forthcoming, 2005). About the Authors Alan Manning is a professor of linguistics and English language at Brigham Young University. He teaches graduate courses in writing and research design and undergraduate courses in semiotics, semantics, and theoretical syntax. He is a coauthor of Revising Professional Writing in Science and Technology, Business, and the Social Sciences (with Riley, Campbell, and Parker, 1999, Parlay Press). Nicole Amare is an assistant professor of technical writing at the University of South Alabama, where she teaches technical writing, editing, stylistics, and grammar. She has written Real Life University, a college success guide, and has edited Global Student Entrepreneurs and Beyond the Lemonade Stand. With Barry Nowlin, she is currently writing Technical Editing: Products and Processes (forthcoming from Prentice Hall in 2006). [7] K. Keller and B. Shwom. The great man has spoken. Now what do I do? : A response to Edward R. Tufte's The Cognitive Style of PowerPoint. Communication Insight, pp.1-16, October Available: InsV1._000.pdf [8] C. S. Peirce, Ideas, stray or stolen about scientific writing, no. 1, (1904, unpublished ms.) published in Philosophy and Rhetoric (1978) 11 (3), pp [9] C. S. Peirce, Collected Papers, volume II, Hartshorne & Weiss, Eds, Harvard Univ. Press, [10] K. St. Amant, personal communication. [11] J. Doumont, "Magical numbers: the sevenplus-or-minus-two myth", IEEE Trans. Prof. Commun. vol. 45, no. 2, pp , June [12] Farkas, David, Explicit structure in print and onscreen documents, Technical Communication Quarterly, vol. 14, no. 1 (forthcoming, 2005). Page numbers cited from online ms. Available: rkas/farkaspub.htm 287