Considered Harmful

I've had several occasions recently to rant about Microsoft PowerPoint, and it's getting pretty tired for all concerned, so I'm hoping to get the whole argument out of my system by writing it down and thereafter never think about it again.

The first thing to point out is that this isn't just some knee-jerk anti-Microsoft stance. I'm certainly no MS fan and need only the flimsiest excuse to bash them mercilessly; but in this case that would be superfluous. When your position is supported by the likes of data presentation guru Edward Tufte and the productivity experience of Sun Microsystems, chances are there's some substance to it.

In fact, PowerPoint is a decent enough piece of software, at least by Microsoft's lights; it's certainly nothing like the farrago of design idiocy and implementation incompetence that is MS Word. Moreover, while PowerPoint is the pre-eminent source and exemplar of the problems I'm railing against here, the same arguments can be levelled more or less verbatim against competitors such as Apple's Keynote; the latter is marginally less ugly, but pretty much as fundamentally egregious.

PowerPoint is absolutely endemic in the scientific community, just as it is in the commercial world. And this is an unmitigated Bad Thing because, in general, PowerPoint is obstructive to good communication. It doesn't have to be -- it is quite possible to construct a great presentation in PowerPoint -- but expressive poverty is its default mode. Instead of enabling better communication, it actively encourages worse. It constitutes an alluring lexicon of banality, awkwardness, tedium and vacuous diversionary chrome. Even great communicators can be unwittingly trammelled by its hapless tropes; and most of us are not great communicators.

Of course, it is silly to expect a piece of software to rescue us from our inarticulacy. If we are clumsy and tongue-tied in front of an audience, PowerPoint isn't going to change that. But it pretends it will. It offers itself as a crutch, which we grab at gratefully; only to compound our problems.

The thing is, especially in a scientific context, audiences rarely expect great oratory. The presentation of science is often dry and bumbling, and that's fine. Delighted as we might be by a barnstorming performance of wit and vivacity, we'll happily make do with interesting results and some kind of clarity of thought. In PowerPoint's false promise of the former we all too often find instead a betrayal of the latter.

Which is all very well, but we appear to be stuck with the wretched thing. How to make the best of such a sub-optimal situation?

There are no end of internet sites offering advice on this topic, ranging from the insightful to the witless. Many of their exhortations are specific to the point of inflexibility, and I have no intention of rehearsing that sort of cookbook approach here. Instead, I'd like to present some more general principles, on the grounds that the whole PowerPoint problem is ultimately that of unthinking. The only way to combat its deadening appeal to the lazy, habitual parts of our minds is to apply the other parts. If we think clearly we need not succumb.

With that preamble -- and with no claim to either authority or originality -- here are my personal PowerPoint Commandments. Read 'em and weep.

1. Do not expect -- or ever allow -- PowerPoint to think for you.

PowerPoint is a drooling village idiot, the very incarnation of dumbed-down, lowest common denominator fatuity in software. Its gormless character must be ruthlessly suppressed; otherwise it will overwhelm your own.

Almost all of PowerPoint's templates, styles, clip art and default behaviours are rubbish. Don't use them, at least not without seriously considering what they provide in the light of what you want to say. Bullet lists are very seldom a good way to get across information. Transitions impart nothing. Why does each slide need a title? Don't fall into these dullardly patterns just because the application happens to.

2. Plan ahead.

While some few gifted souls may be able to conjure a lively presentation ex tempore, chances are you aren't one of them. You need to work out what you're going to say, and how, before you begin. Not only before speaking, but also before creating your slides.

Don't write your presentation in PowerPoint. If you insist on using PP as an outliner, do so in a different file from your actual slides; better, use a pen and paper. Construct a cogent line of argument. Define your points. How are you going to get from A to B to C? Determine the data and images needed to back the points up. Rehearse your presentation without slides.

In short, think. I know, I said that already. It bears repeating.

3. Your slides are not your script.

Never read the text off your slides. Never, ever, under any circumstances. Don't even use slides as cues if you can help it. Not only does it give you no communicative benefit at all (see rule 4), it actively undermines what you're saying. Your audience can read faster than you can talk. If the slides are telling them what you're going to say, your whole presentation becomes an exercise in going through the motions. You might as well just shut the fuck up.

The same goes for giving out handouts of your full text beforehand. People will read them instead of listening, which means they won't even be looking at you or your slides.

If you need a script or a list of cues -- and, let's face it, you probably do -- put them on a piece of paper or an unseen computer screen. It is perfectly possible to read from a script and seem spontaneous and interesting, especially if you have enthusiasm for your subject, but if the script is right there for everyone to read in foot-high letters you can only ever come across as a mindless automaton.

4. Exploit the available channels.

That you are using PowerPoint at all means you have two distinct channels of communication to your audience, visual as well as verbal. Make use of them both, don't let one be a weak facsimile of the other. Show and tell.

Your slides should back up, illustrate, emphasise, decorate or counterpoint what you are saying. They should show the data or capture the essence in some way that your spoken words alone do not. If they add nothing to the experience, why bother with them at all?

5. Less is more.

This is not hard and fast, but in general you should aim for simplicity in your slides. In particular, one slide should illustrate one point. This does not mean that you need to have a slide for every point; some points may very well be better off unillustrated. It does mean that you shouldn't try to cram lots of points onto a single slide. The purpose is to communicate, not to obfuscate.

Don't simplify to the point of dimwittedness. If you are illustrating a complex point or presenting complex data, you may need to have some complexity in your slides. But don't include things that don't contribute out of some sense of duty or desperation or because you hope they'll confer an unearned aura of legitimacy. They won't.

6. Use the right tool for the job.

PowerPoint includes basic tools for image adjustment, drawing and data presentation, but they are rubbish and should be avoided for anything other than the simplest graphs. Prepare your data in R or Mathematica or even Excel. Edit images in Photoshop or the Gimp. Create diagrams in Illustrator or Inkscape or Visio. Set equations in LaTeX. Hell, you can even write and draw stuff on paper and scan it in.

There are excellent tools available for everything you'll want to achieve, many of them free and open source. Don't fall back on PowerPoint's substandard versions just because they're there.

Make sure the input files are in widely compatible formats; PowerPoint has a nasty habit of using platform-specific mechanisms to render some image types, leaving empty boxes when playing back on other machines. You're probably safe with Microsoft's own BMP; steer clear of TIFF. Downsample images to a reasonable size before importing; resizing huge pictures at showtime can lead to all sorts of grief.

Only collate everything and lay it out in PowerPoint once it's all in an appropriate form. And pay some attention to the layout and visual organisation. Does the slide look good? Is it clear or an incoherent patchwork? Nobody expects you to be Neville Brody or Dave McKean, but it helps if we at least get the impression that you care about your content.

7. Know your stuff.

Planning only gets you so far. If you're going to be able to respond to questions or work around technical hitches, you need to know your subject inside and out. Have your line of argument at your fingertips. Understand its strengths and weaknesses. Be familiar with the data. If there are particularly shaky points, acknowledge them.

A thorough knowledge of what you're talking about allows you to be flexible. Have points that you can throw in or leave out as seems appropriate. It frees you from dependence on your slides. It will give you the confidence to project better.

You might still be a babbling inarticulate buffoon despite being in complete command of your subject, but at some level that mastery will come across. Better that than a glib, smarmy salesdroid with a smile full of orthodontistry and a brilliantined tongue, who can read his slides with dizzying fluency but doesn't really understand a word.
Posted by matt at April 20, 2007 02:03 PM