Thoreau offers without qualification some complaints about a paper in a glamour journal, ending with:
All of this might have been excusable if the big flashy Glamour Journal paper had been followed up with more detailed papers in other places (a common practice in some fields). However, when I searched to see what the authors have done since and whether they’ve cited that paper, the only places I found them citing their own paper was in papers only marginally related to the work published in the glitzy place. So there was no follow-up, just something that was trendy enough to get into a flashy place.
This is a problem that has existed for years– I remember doing literature searches on paper in my undergrad days, trying in vain to locate the forthcoming paper-with-technical-details promised in some journal article. Which makes this a good topic for a poll, because I bet Thoreau and I aren’t the only ones who have had this experience:
I have a bunch more to say about this, but it won’t fit within the page limit of this blog post, so look for a forthcoming article on the subject Real Soon Now.
This is part of why I think Science and Nature are harmful to science…. (The main reason is the whole “proprietary” thing.)
However, you can get bitten by not throwing out the glamour paper. In January 1998, as evidenced by our posters at the AAS meeting, my group (the Perlmutter group) was ahead of the other group (the Schmidt group) with data showing that the expansion of the Universe is accelerating. This lit a fire under them, while we spent the next 8-9 months being very careful about doing all the cross analyses and redoing everything and so forth to produce the best technical paper we could. The results was Riess et al., published March 1998 (http://adsabs.harvard.edu/abs/1998AJ….116.1009R) and Perlmutter et al., published in early 1999 (http://adsabs.harvard.edu/abs/1999ApJ…517..565P).
Historians looking back will see that the Schmidt team published the results first… although if they read the introductions to the papers, they’ll see that both teams like to claim that they were first. (The only rational interpretation nowadays is that both teams did it at the same time.) (There was a bunch of internal politics with the Knop 2003 paper about how we were going to state the 1998/1999 results, to represent the reality of publication without ceding priority for our group… sigh.)
The thing is, if we’d written a “glamour” paper, we could have had the results out “pending a forthcoming longer paper” in February 1998. It wouldn’t even have substantially delayed or changed the longer paper that was submitted in September 1998. Or, at least, it wouldn’t have delayed it for reasons of our time; it might have taken the pressure off and allowed us to dither more….
Rob @#1 gives one example that argues (correctly IMO) for putting out the letter ASAP, but IME it is not uncommon that the long paper never comes gets done. It might live in a PhD dissertation, if one was associated with the project and you know who to look for, or it might never get written in any form. It seems worst in a rapidly advancing field.
Should repeat offenders be held accountable?
This same behaviour goes on in lower prestige journals. It is just less obvious because of the overall lower impact.
In contrast, in my field I see a lot more weight given to publications with large public data sets. It is not that the results are more original or unique, in fact often I see complaints about rediscovering the wheel. But, I think that these results get around exactly the problems raised by Chad and Thoreau.
Equally bad is citing the critical paper as “in press” when it isn’t. Caught this in a manuscript I reviewed recently, an “in press” article by one of the senior authors. But when an earlier published article was checked for more information, the same “in press” citation was used in that paper, and it was published 4 years ago. In press for 5 years and counting, sure. Really pissed me off, and worse, it’s a cheat, a way to avoid answering some tough questions.
Yes, I have run into this issue, or variants thereof, over the years. It’s not just GlamourPubs either, although the GlamourMags are the worst at this kind of thing. More than once I have run across a reference in a paper to an unpublished manuscript by Fulano et al., and I’ve had to ping Dr. Fulano (or in some cases his advisor, since some people do leave the field after getting a Ph.D.) for help tracking down the manuscript in question. Other times there is a citation to a meeting abstract (some journals in my subfield allow this practice), and the reason people cite the abstract is because the work never got turned into a published paper. There is also at least one case in my field where a manuscript remains unpublished because it was in the author’s checked bag which the airline temporarily lost (this was back in the 1970s when all manuscripts were hard copy) and he had moved on to other problems by the time the bag was returned to him six months later.
In Mathematics the most prestigious journals publish long articles. Forty pages is average, above 100 not uncommon, 200+ rare but possible. Because good mathematics is forever, or at least for a few decades, and i’s worth a detailed explanation.
Of course we don’t publish in Glamour Journals, but I think this is because we never have colour pictures. Often no picture at all, since it’s very hard to picture something which is 20 dimensional, or negative dimensional.