Suggestions for Delivering Invited
（APS Division of Nuclear Physics）
アメリカの物理学会の Nuclear Physics
The following suggestions are designed to help invited speakers at Divisional meetings to prepare high quality talks. They are modeled after a paper written by K.K. Darrow and published in the February, 1951 issue of Physics Today, and reprinted there in the December, 1981 issue. Some of these suggestions will no doubt be considered by some speakers to be the painful elaboration of the obvious. However, we have all suffered through talks in which the obvious has been ignored.
Specific suggestions and aids to preparation
Speak loudly enough to be heard at the back of the room. The acoustical conditions and quality of amplification available will inevitably be quite variable from meeting to meeting. It is thus important to direct one's remarks to the back of the audience. In particular, the common mistake of peaking in a low voice to (possibly) distinguished persons seated in the first few rows should be avoided.
Speak to the audience. It has now become quite common to use transparencies in place of written notes; this probably improves the quality of most talks. However, it is important to avoid the mode of facing the screen and reading from your transparencies. Maintaining eye contact with the audience is an extremely effective way of improving the quality of a presentation, and should be consciously attempted wherever possible.
Introduce your subject by placing it within the general context of nuclear physics, or, if possible in a manageable amount of time, physics as a whole. The typical audience for an invited talk is quite diverse, and will in general consist of people in different subfields of nuclear physics who are interested in learning what is new and interesting in the subject of your talk. Time devoted to relating your specific subject to current issues in nuclear physics can make the difference between most of these people getting something out of your talk or not. Also, an invited talk should consciously be aimed at the average member of the audience, not the 3 or 4 specialists who are doing closely related work.
Practice the talk, including projecting the transparencies and timing yourself. It is difficult to overemphasize the importance of this step. By doing this you: 1) eliminate rough edge, clumsy language, implausible lines of argument, etc.; 2) ensure logical flow, and find out which transparencies "work" and which don't; 3) get the time right, and make sure that there is time at the end to summarize the conclusions without rushing. This last point is especially important, especially for any speaker who prepares his or her talk by modifying a talk originally designed for a 1 hour presentation. (In such cases it is almost always necessary to decide at the beginning to restrict the scope of the talk in some way in order to make the length manageable.)
Pay careful attention to preparing transparencies. Probably the best general rule is "keep the information density relatively low." Lettering should be large enough to be read at the back of the room. Use of colors can be very helpful to add interest and clarity, but remember that contrasting colors are easier to see, and that some marker pen colors (e.g. yellow) can be nearly invisible. Multiple graphs and complicated typewritten tables should be avoided wherever possible. If it does prove necessary to show a complex set of equations or set of experimental data the presentation should be carefully prepared, and visual aids should be used to draw the audience's attention to the important points. The overall number of transparencies should also be kept to some reasonable limit, say 15 or 20 for 30-minute talk. It is also important to pay attention to the organization of the information on the transparencies. With a little care it is almost always possible to present one transparency at a time in such a way that the information you need is on the screen when you want it. The practice of gradually uncovering a transparency as you speak should usually be avoided. In cases where it is essential to insert information in the middle of a transparency while discussing it, overlays can be helpful. The general principle of keeping it simple should be particularly kept in mind when using overlays, however. Also, remember not to block the projection by pointing to the transparency on the projector. Use a pointer and point to the projection screen instead.
The most important single thing to do in preparing an invited talk is to imagine yourself in the audience with no very specialized knowledge of the subject -- and then to avoid boring or confusing yourself too much. We have all heard talks ranging from the very good to the utterly abysmal. It is clearly in all of our interests to make the quality of presentation of Division of Nuclear Physics invited talks as high as possible.
In preparing these remarks we have attempted to update and adapt the original guidelines in Darrow's article to the conditions prevalent today (i.e. transparencies, etc.). However, in our opinion it is well worth reading the original article. It certainly captures in spirit, if not in contemporary detail, the ideal of what an invited talk should be.
Final Revision November 4, 1986