SUBJECT: How NOT to Give Talks

Presenting data at a conference? Preparing a seminar or lecture? Then you need the Immunology News "Guidelines For Giving a Truly Terrible Talk." Strict adherence to the following time-tested guidelines will ensure that both you and your work remain obscure and will guarantee an audience of minimum size at your next talk.


1. Use lots of slides and overheads. A rule of thumb is one for each 10 seconds of time allotted for your talk. If you don't have enough, borrow the rest from the previous speaker, or cycle back and forth between slides and overheads.

2. Put as much information on each slide and overhead as possible. Graphs with a dozen or so crossing lines, tables with at least 100 entries, and maps with 20 or 30 units are especially effective; but equations, particularly if they contain at least 15 terms and 20 variables, are almost as good. A high density of detailed and marginally relevant data usually preempts penetrating questions from the audience.

3. Use small print. Anyone who has not had the foresight to either sit in the front row or bring a set of binoculars is probably not smart enough to understand your talk anyway.

4. Use figures and tables directly from publications. They will help you accomplish goals 2 and 3 above and minimize the amount of preparation for the talk. If you haven't published the work, use illustrations from an old publication. Only a few people in the audience will notice anyway.

5. Make sure at least one slide and overhead is upside down or sideways. This relieves tension in the room.


1. Don't organize your talk in advance. It is usually best not even to think about it until your name has been announced by the session chair. Above all, don't write the talk out, for it may fall into enemy hands.

[Editor's Note: Actually, if you want to give a *truly memorable* presentation, WRITE OUT THE ENTIRE TALK word for word -- and read from the script in as close to a monotone voice as you manage. This method is especially effective right after lunch.]

2. Never, ever, rehearse, even briefly. Talks are best when they are given spontaneously with thoughts organized in a random fashion. Leave it as an exercise for the listener to assemble your thoughts properly and make some sense out of what you say.

3. Discuss each slide and overhead in complete detail, especially those parts irrelevant to the main points of your talk. If you suspect that there is anyone in the audience who is not asleep, return to a previous slide and discuss it again.

4. Face the projection screen, mumble, and talk as fast as possible, especially while making important points. An alternate strategy is to speak very slowly, leave every other sentence uncompleted, and punctuate each thought with "ahhh," "uhhh," or something equally informative.

5. Wave the light pointer around the room, or at least move the beam rapidly about the slide image in small circles. If this is done properly, it will make 50% of the people in the front three rows (and those with binoculars) sick.

6. Use up all of your allotted time and at least half, if not all, of the next speaker's. This avoids foolish and annoying questions and forces the chairman to cut short the following speaker's time. Remember, the rest of the speakers don't have anything important to say anyway. If they had, they would have been assigned times earlier than yours. · · Homepage
Last updated 20 Jan 97