by Rachel A. Russell, American Mineralogist Managing Editor

I hope the dramatic changes in production methods at the American Mineralogist are completely invisible to the average reader. But I'm quite sure the average author has been surprised to see on the proofs that some (or possibly many) symbols and special characters haven't translated correctly. The ultimate solution will be for improved technology, and I have no doubt that's around the corner. But in the meantime, what can you submit as final revision that will have a greater chance of translating correctly?

First off, note that none of this matters in terms of what you submit. Make what you submit as readable for the reviewers as possible. Concentrate on the science you trying to explain or study. But when things seem to be progressing happily and you’ve been asked to supply an electronic revised copy, then things get serious around here! However, I don’t want the authors to be burdened with "codes" or formatting, and in fact, the less formatting the authors do, the better. Formatting is our job. Still, there are a few things authors can do that will produce a cleaner proof and thus give all of us an easier task.

The first thing to do is to use "symbol font" for symbol characters such as alpha, beta, chi, gamma, mu, and so on. So instead of using Word's special character box, type a regular a, highlight it, and change it to the symbol font. An alpha should appear on your screen (and print). I've actually made keyboard shortcuts on my Mac that let me do this quickly. However I have to find "by hand" each instance of the non-symbol symbol in order to convert it and that's why we miss some. Auto-finding, or searching and replacing, doesn't usually work, but sometimes it does work. That's why we all love computers!

For special symbols such as arrows and square root signs, I think there is no alternative to using the special character box. We must all be eagle-eyed in catching the mistranslations on the proof. Authors should keep especially alert for these symbols when reviewing their proofs. Every now and then, instead of mistranslating by inserting a nonsense character, it will mistranslate by putting nothing there at all, not even an extra space. That’s particularly tricky to detect. The other computer trick is to insert a wrong symbol that’s not complete nonsense at first glance. For example, we’ll see 10% and think nothing is wrong, whereas the author will know that should be 10°.

For some reason, the computer’s favorite thing to mistranslate are degree signs. It helps when the author uses the symbol for a degree sign, in Word on a Mac that is opt-shift-8. It is surprisingly time-consuming when the authors use superscripted or raised zeros or "ohs". Search-and-replace methods have been very helpful in this situation.

However, there are two cases where authors often go to too much trouble to "do it right". The first case is that of "stacked" characters. Authors can just superscript the supers and subscripts the subs, they do not need to actually stack them in Word. We do the stacking in PageMaker itself because any Word stacking ends up lost. Whatever you do, please do not use an equation editor to create simple in-line stacks! We have to retype each one of these. The second case of "working too hard" is that of overbars. We create the overbars in PageMaker, not Word, so each instance of overbars in Word or in an in-line equation box has to be deleted. We type 3bar, for example, to mean an overbar goes over the 3. Authors could simply jot a line over the 3 on the hardcopy, or put a minus/hyphen sign in front, or even a note on the first instance to say some like "in all instances of P3, the 3 has an overbar." On the other hand, if it is easier or more comfortable for the authors to continue inserting the overbars, that’s pretty understandable and we’ll just work it out here.

I guess my final plea is that authors use superscripts and subscripts and NOT raised or lowered type!