The invited review ? or, my field, from my standpoint, written by me using only my data and my ideas, and citing only my publications.

Do you feel scientifically isolated? Do you find yourself sitting on the side-line while others take the field by the nose and lead it? Are you unable to publish a model that summarizes your data and ideas because reviewers label it as being too speculative and unsupported? Can't get those experiments published in any regular journal? Do you find that nobody is citing your papers? Haven't published in your field for some time, but want to show that you are still a player? Well, no need to worry! There is a special category of publication for you, 'the invited review', and even specialized journals, the 'review journals', that cater to every part of your neurosis. The major difference between many invited reviews and regular articles in journals ? this journal being an exception ? is that the invited review is generally not critiqued by outside, independent referees. In those rare cases when a review is critiqued, the letter from the editorial staff that accompanies the manuscript usually contains a disclaimer: "While we are interested in your critical evaluation of this manuscript, you should know that this review was solicited by the Editorial Board, and, therefore, we allow some latitude in the content and opinion expressed." What's the point of reviewing it when the writer has been given some sort of carte blanche to say what they want? Are you surprised? I have critiqued reviews to no affect (but then that's true also for research papers!), suggested that the writing was too biased, that the appropriate references were not cited and that the proposed model must have been conjured up while the writer was inebriated. Of course, all the critiques for my reviews were taken seriously, and I made all the suggested changes ? right! So, what kinds of review are there? As a postdoc, I got introduced to two extremes in one conversation. I was asked to write a review by my mentor. Not being very opinionated (at the time), I read hundreds of papers and cited them, and wrote a discussion of the results without providing a, how shall I say, personal opinion. My mentor scanned the manuscript, and said: "Great! Now start again. This time, write a review that contains lots of ideas and models, minimal referencing, and by all means speculate wildly on where we think the field is moving. Oh, don't throw away this one. We can use it another time for one of those 'Annual Reviews' journals." Great mentoring, huh? So, what's the point of an invited review? Well, for the lazy ? sorry I meant busy ? it is a way to keep up with publications, ideas, trends, and models in a specific field. But of course, this is exactly where the problem lies. Unless the reviews are balanced (i.e. unbiased), complete in citing publications (i.e. not just the author's papers), scholarly in the presentation of models (i.e. present competing models), what sort of intelligent way is this to 'keep up with a field'? What we need is a more critical approach to reading published papers. That's right, we should discuss the fact that a certain experiment doesn't make sense, a control was missing, and an interpretation based on the data presented is crazy. Ah, but that would be so impolitic, wouldn't it! I have two suggestions. First, let's get people to write reviews outside their own specific field of interest. There are many advantages to this. There would not be a bias, or a limited interpretation based on 'reading between the lines' or on unsupported experiments. The review would be a critical evaluation of data presented (only) in peer-reviewed research papers available to the average reader and comprise a balanced discussion of all the data from competing labs. There is of course a major disadvantage: who has the time or gumption to do this (just wait for the reprisals)? Well, most of us participate in some sort of journal club in which papers outside our field are presented. Perhaps, these discussions should be published (on the Internet?). [Some journals provide 'sound-bites' of papers (Paper Alert, or some such heading), but these tend to regurgitate the paper's abstract and are hardly critical.] Second, without implementation of the first suggestion, I think that we should add a warning sign to invited reviews: "WARNING. THE CONTENTS OF THIS REVIEW MAY BE DANGEROUS TO YOUR SCIENCE. THIS REVIEW CONTAINS STATEMENTS, IDEAS AND OPINIONS THAT MAY BE UNSUBSTANTIATED. THE MODELS HEREIN REFLECT THE STATE OF THE AUTHOR'S LOOSE GRASP OF REALITY, AND SHOULD NOT BE CONFUSED WITH FACTS. PROCEED WITH CAUTION." Invited reviews are no substitute for reading the primary literature and critically forming your own opinion. An occasional column, in which Caveman and other troglodytes involved in cell science emerge to share their views on various aspects of life-science research. Messages for Caveman and other contributors can be left at [email protected]. Any correspondence may be published in forthcoming issues. Previous Sticky Wickets can be viewed at: www.biologists.com/JCS/caveman/index.html jcs6024