david farkas on "how powerpoint adversely mediates thought"


I'm at an MSR talk by David Farkas from UW entitled "Need: How PowerPoint Adversely Mediates Thought and Possible Remedies." Since anyone who's been reading my blog for a while knows how much I dislike most uses of PowerPoint, I'm particularly keen to hear why Farkas thinks Tufte is wrong, and also what he suggests as remediation.

He cites a Microsoft estimate of 30 million powerpoint presentations being given per day. Ouch. I wonder if that's a verifiable statistic.

(In an aside, Farkas notes that PPT is more constraining than Word, since Word provides more of a blank canvas. Farkas says that Tufte believes Word has no cognitive style, while Farkas says that's obviously not true--our tools inevitably shape our message. I'd have to agree with that.)

Discusses a critique by Ian Parker called Absolute Powerpoint, as well as Tufte's The Cognitive Style of Powerpoint.

Where Tufte claims that PPT encourages deep hierarchies (many levels of nested bullets), Farkas argues that the reverse is actually true--there's an upward vector on content, resulting in a flattening of hierarchies.

Farkas asks us to take a step back and look at the larger picture of presentation contexts--audiences, presentation genres (product rollout vs technical briefing vs slide show). Tufte focuses on technical genres of presentations, whereas many PPT presentations are focused on "light" genres.

He says you can't assess a PPT deck outside of the context of the presenter's performance/style. The amount of time spent on each slide, whether or not there's a handout provided, and other factors can influence the effectiveness of the deck. The audience has to be considered, as well...their information needs, their cognitive styles, all impact the effectiveness of the presentation.

He describes PPT as inherently "topo-centric"--the presentation of each slide is static and fixed, rather than scrolling. This is good because it provides persistent context during the presenter's discussio of the slide, but bad because it flattens hierarchies.

Print (and HTML), Farkas argues, have a "downward vector and a nesting problem." Print hierarchies naturally run deeper. (Missed the rest of this because I was briefly distracted...)

"The PowerPoint distortion hypothesis" - It is highly plausible that PPT causes deck authors to distort the visual representation of their logical hierarchies. What, then, are the implications for audiences and presenters?

He uses an example deck to show some of the distortions that can occur, but I'm unable to see anything but the ugliness of this deck. White and yellow Times text, in seemingly random sizes, on a bright purple background. Why does discussion of content always seem to ignore the impact of aesthetics?

Oneof my MSR colleagues questions the underlying assumption that all content is hierarchical. Farkas argues that this is necessarily true, that it's a function of how we think. I'm not convinced--many of the best powerpoint-supported presentations I've seen used no bullet points, and no explicit hierarchies.

When I raise my concerns, he responds by saying he wants to limit his discussion to the genre of presentations that need to present hierarchical content--main ideas, sub ideas, supporting material, etc. I'm still not convinced. One of my frustrations with PowerPoint is that it does in fact push that idea on us--that presentations are and should be made up of hierarchical point/subpoint content. In fact, the people doing the best work with PPT tend to go "beyond bullet points," and use it as a narrative medium. But that doesn't prevent them from presenting very detailed and even technical information--it's just that they're presenting it in a way that doesn't fit into this hierachical structure. (For example...Dick Hardt's identity presentation, or Lawrence Lessig's inimitable talks.)

Another commenter argues that the slide should be the secondary channel, and the presenter should be the primary channel. I wish more people here thought this way...that's a big part of what I was critiquing when I wrote about the "culture of the deck" here.

He makes a number of suggestions about how to make it possible to show complex hierarchies more easily in Powerpoint...something that, quite honestly, makes me cringe. I do like his suggestion, however, that you provide breadcrumb-like information at the top of each slide to show where it fits in a hierarchy (if you choose to make your hierarchy explicit).

Suggests some good directions for future research. How do audiences process information in presentations? Can we better udnerstand deck authoring processes? And the last, which I find last compelling, "develop a meaningful taxonomy /vocabulary of deck content and glossing behavior." I'm not sure we need special language to describe ideas/content in decks as opposed to other text or graphical materials.

An audience member I don't know points out the extent to which the slides were forcing him as a presenter to stick to a script, and not engage the audience. He notes that Farkas engaged the audience in his presentation exactly 3 times, and never more than for ten seconds at a time.) Is this an effect of the hierarchical structuring of the content?

Another questioner asks about ways that we can support more creative presentation styles and more creative presenters. The research question here is "are there ways to identify, incorporate and disseminate best practices in presentation methods?"

There's some discussion about the prevalence (seen as both inevitable and necessary) of PPT decks as standalone documents as opposed to presentation aids. Also some discussion about the ways that the "Notes" section can be used in that context.

(It's amazing to be in a room full of not just smart researchers, but also the people who actually build these tools...several PowerPoint team members are here, responding directly on the intended use of specific features. I will so miss this about MSR talks.)

Farkas slams Atkinson's "Beyond Bullet Points," describing it as destructive rather than helpful. It's a terrible direction, he says, to throw out bullet points entirely. It's throwing out the baby with the bathwater.

There's some discussion then about Atkinson's argument, which Farkas says is based on Richard Mayer's work, but is a distortion of it. (Just looked at Mayer's web site, and his work looks fascinating. Note to self: bookmark that for summer reading.) Is it a problem if you've got material on the screen that's unrelated to what you're talking about?

All in all, I'm left with more questions than answers (which isn't necessarily a bad thing), and a sense that there needs to be a great deal more research into the underlying assumptions on presentation methods and materials.


Bonus links to PPTs I've seen and loved:


It is a mistake to consider powerpoint as a stand alone information vehicle.

It should be a supporting actor in a performance. I am not sure how many presenters consider presentation as performance, but they are. (The excellent examples you provide showcase PERFORMERS who use powerpoint).

Think mnemonic - not meat.

There's a lot to disagree with in Farkas' talk. Slide decks as handouts are not inevitable. It's way more than evaluating the slides in the context of the presenter. The presenter is the presentation. The slides support the presenter.

Presentation Zen is my favorite site about presentations and presenting.

I love Presentation Zen, too.

But the reality in corporate America--not just at Microsoft, either--is that PPT often *is* used to create stand-alone deliverables. Saying that's not good doesn't make it stop. :) So the question is what do you do about that reality? How do you design the tool, for example, so that it gives you different "autocontent" suggestions if you tell it you're creating a presentation aid vs a document to be distributed?

People use PPT for non-presentation content because (in part) it allows them to (a) avoid writing, which many people aren't good at, and (b) chunk information up and illustrate it in a way that tends to work well for busy managers.

It is true I also use powerpoint as a prototyping tool. It is an easy way to make a series of still with hot spots that can be linked between slides. In that case, it is not a presentation - but a low fidelity interaction prototype.

Within the notes, I can talk about what is happening under the hood, place pertinent questions related to the slide, and elicit input for the specifc page.

The notes are my stand in, walking others through the show. Still...the content of the page ends up being devoid of minutinae. Perhaps the strategy would be to allow for more information to reside in the notes area - and disallow more than illustration in the slide?


I agree that the challenge is not to present a better hierarchy. It is to present a better flow. Powerpoint does tend to default to the bulleted list. For a new deck, the program prompts the user with a set of options (the default being the title slide with a title and subtitle zone; the next being the bulleted list and then variations of column and table layouts; the last is the blank slide).

I wonder if users who were exposed to Hypercard approached Powerpoint for its ability to navigate different sequences within the same presentation. I ask because "slide" in Powerpoint is very much connected to the image of "page" in the mind of the user [an effect due largely in part to its being bundled in an Office suite of products with a page-centred (Word) as opposed to document-centred (Wordperfect) wordprocessing software].

The very language of "cards" in Hypercard encouraged the notion of shuffling (by analogy with the experience of index cards). "Slides" are slotted in a carousel or tray. Hence I would suggest that it is more about how programs through the interface metaphors offer most users a default approach and other users a chance to resist the default.

BTW, the title of David Farkas's talk is difficult to parse. "Need: How PowerPoint Adversely Mediates Thought and Possible Remedies." Remedies to thought? Mediation between Thought and Possible Remedies? The very title is caught in the amiguities of reading a hierarchy from the bottom up.

Leave a comment

About this Entry

This page contains a single entry published on June 21, 2006 1:34 PM.

stupid error messages was the previous entry in this blog.

simple (but not easy) advice is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.


Category Archives