Table of Contents Table of Contents
Previous Page  3 / 68 Next Page
Information
Show Menu
Previous Page 3 / 68 Next Page
Page Background

Lab Times

5-2014

page

3

Editorial

Do We Still Read Each Other’s Papers?

... This question is obviously as legitimate as never before

these days.

A couple of years ago, for example, we asked Axel Brennicke,

Chair of Molecular Botany at the Universty of Ulm, what exactly

his motivation was to regularly provide our German sister journal

Laborjournal

with his meanwhile very popular, sharp-tongued

op-ed pieces (see

www.laborjournal.de/rubric/archiv/domfac

/

ansicht.lasso

; German language!). Of course, he had a number of

good reasons with regards to content but he finally also added,

“And, by the way, those pieces are by far my most-read ‘papers’”.

We asked him to take a guess at how many people he thought

would really read his

true

research papers. “Not counting the peo-

ple in the lab, I think three to five,” he answered with a grin.

Okay, that was meant as an

understatement. However, let’s

not fool ourselves – it is very

likely that the true numbers are

not that significantly higher.

Just take a look at a normal, av-

erage and solid “middle-class”

paper. Who would read it? Of

course, there would be three to

five editors and reviewers who

have to

read the manuscript

ex

officio

; furthermore, a couple

of people from the own group

as well as five to ten “friends” or cooperators join in – and, in ad-

dition, possibly the same number of competitors. And if you are

very lucky, yet one or the other grad student, for whatever rea-

son, chooses yours of all papers for his Journal Club talk. So that

makes about 20-40 “readers” for the “average paper” in the end.

An estimate still too optimistic? Well, probably. Claus Wilke,

computational biologist at the University of Texas in Austin, for

example, recently also dealt with this topic in his Blog

The Seri-

al Mentor

. At the end of his “analysis” he came to the conclusion

that less than ten percent of his papers are read by more than ten

people (with his “best” paper having been cited almost 500 times,

after all). No wonder, therefore, that he provocatively titled his

essay “No one reads your paper either”.

Do those dark number scenarios come as a surprise?

Just remember how it was about twenty years ago. Whenever

you went to your institute’s library you met dozens of people there

reading articles

. More than a few of them had even reserved one

fixed half-day per week just for this task: screening the current lit-

erature and

reading

(and sometimes xeroxing) interesting articles.

Today, in the times of internet and databases, following the

current literature and tracking down relevant articles has become

a whole lot easier – even with many more papers published. Given

this, we must state: Yes, it does come as a surprise that researchers

obviously read less papers today than 25 years ago.

In the recent “20 Years Anniversary Issue” of

Laborjournal

,

Diethard Tautz, Director at the Max-Planck-Institute for Evolu-

tionary Biology in Plön, also bitingly wrote about the issue. The

trigger was that Tautz originally had analysed, in which way cer-

tain articles from his group were cited by others. His final conclu-

sion was disillusioning, “In the majority of cases we were cited in

a context, which clearly suggested that the authors couldn’t have

read more than just the title of our paper.”

However, this is not really hot news. Already in 2003, the US-

mathematicians Mikhail Simkin and Vwani Roychowdhury tried

to estimate “what percentage of people who cited a paper had ac-

tually read it”. Their method of choice was to analyse the dissem-

ination of misprint occurrences in citations and to test, stochas-

tically, whether those misprints might mainly spread through

purely copying the reference from the reference lists of earli-

er “mistypers” – without ever

reading

the corresponding paper.

Their final conclusion after pages of plots and formulas, “Our es-

timate is that only about 20% of citers read the origi-

nal” (

Complex Systems

14: 269-74).

All that leaves one question in the end: If even the

people who cite your paper don’t read it – who else is

actually likely to?

For Tautz, all this is already sufficient evidence to

envision an alarming development: “At the heart of

science is to ask new questions that build on existing

knowledge. Building on the knowledge of others, how-

ever, increasingly melts away – just because we do not

read their work any longer.”

And with the last anecdote he presented in his es-

say, Tautz even went one step further. “Recently,” he

wrote, “I had to review one of those ‘big data’ papers including 60

authors and a firework of intelligent analyses, statistics and sim-

ulations. The strange thing, however, was its modular construc-

tion. Apparently, the paper was written by various authors, each

one being a specialist for one field with own writing style and lit-

tle relation to the other modules. Some odds and ends, howev-

er, like lacking transitions and incomplete sentences, clearly in-

dicated that obviously none of the 60 authors had ever read the

complete paper from end to end. Otherwise, at least one of them

should have noticed those flaws. At the same time, I as a review-

er was also only able to evaluate a certain part of the paper; the

remaining parts I could merely check for plausibility. And I guess

the other reviewers felt just the same. The paper was finally pub-

lished; however, I have the strong suspicion that there is not one

reader out there any more who truly understands it

completely

.”

No wonder, Tautz finally added the next, even more disturb-

ing question to the opening one in the title: Do we still

under-

stand

each other?

Photo: openstack.org