Managing scholarly reading

Reading, after a certain age, diverts the mind too much from its creative pursuits. Any man who reads too much and uses his own brain too little falls into lazy habits of thinking. —ALBERT EINSTEIN

How much literature should one read as an academic? Of course, the answer will vary by field, but even within my own field, I find little consensus as to the "right" amount of reading to do.

It is true that no one can read everything that is published, even in a single field such as cognitive science, while maintaining one's own productivity. In my Google reader, I subscribe to the RSS of 26 journals, and from these, I get an average of 37 articles per day. However, in an average day, I feel like I should pay attention to 5 of these. If I were to closely read all of these, I would run out of time to create new experiments, analyze data and write my own papers.

It turns out that in an average day, I'll read one of these papers and "tag" the other 4 as things I should read. But this strategy gets out of control quickly. In May, I went to a conference, didn't check my  reader for a couple of days and came back to over 500 journal articles, or around 35 that I felt deserved to be read. I have over 1300 items tagged "to read" in my Zotero library. At my current rate of reading, it would take me over 3.5 years to get through the backlog even if I didn't add a single article to the queue.

So, how to stay informed in an age of information overload? It seems that there are a few strategies:

1. Read for, rather than read to. In other words, read when knowledge on a particular topic is to be used in a paper or grant review, but don't read anything without a specific purpose for that information. According to proponents of this method, information obtained when reading-for-reading's-take will be lost anyway, leading to re-reading when one needs the information.

This method vastly decreases the overwhelming nature of the information, and makes info acquisition efficient. However, it is not always practical for science: if you're only reading for your own productivity, you're going to miss critical papers, and at worst, are going to be doing experiments that were already done.

2. Social "reading", augmented by abstract skimming. In this method, one does not spend time reading, but spends time going to as many talks and conferences as possible, learning about literature by using the knowledge of one's colleagues. This method seems to work best in crowded fields. The more unique your research program, the more you'll have to do your own reading. And all of this traveling is time and money consuming.

3.  Don't worry about checking through many journals, but set alerts for the specific topics. My favorite is PubCrawler, suggested by Neuroskeptic. Works well when my key words and the authors' key words coincide, but I seem to have set too many topics and I get both too many "misses" and "false alarms".

How do you keep up with literature?