"Standing on giant's shoulders" is a common metaphor for scientific progress. In order to be a scientist, one must first climb the body of the giant, i.e. the accumulated knowledge of previous generations. Reading the published work of other scientists is therefore the most fundamental activity that we perform as academics. The Internet is changing not just the way we use the giant, but also how the giant grows with the accretion of new knowledge.
There are two ways in which scientists learn about relevant literature. One is to browse new publications, another is when they get cited by other papers. The former is more common in fast moving fields like medicine and physics, but the second is widespread in my own field of ecology, where the longevity of most research papers (judged by the half-life of citation decay) is in excess of a decade. The Internet has far-reaching consequences for both modes of knowledge acquisition.
Reading new publications has been revolutionised by services that alert us via email whenever new papers are published in a defined topic area. This means it's no longer necessary to spend time in the library looking though tables of contents (TOC). Although this has obvious benefits in efficiency, there is a cost in terms of the breadth of articles we are likely to consume. In the old days, one would glance at all the titles and perhaps most of the abstracts in a particular journal issue. For example, the current issue of the journal Ecology contains articles on bacteria, plants, insects, fish and birds, covering a wide range of research topics, both theoretical and empirical. Electronic TOC alerts mean that most researchers encounter only articles in their own area of specialism and are therefore much less likely to come across new and potentially transformative ideas. There is a paradox here: the Internet offers the potential to access the full spectrum of research papers, but actually results in a narrowing of focus. The same phenomenon has been observed in online social networks, which are no more socially and ethnically heterogeneous than real ones.
The Internet revolution has equally profound consequences for the second mode of knowledge acquisition. In the old days, I would read an article from start to finish and make a list of relevant citations to fetch from the library. Nowadays, the ubiquity of electronic articles in portable document format (PDFs) means I can get the cited article on screen in just a few clicks. There's no longer any need to move from my desk, or even to finish one article before going on to the next. Often when reading a PDF, I simply scan the text in search of a key assertion or statement. This changes the very nature of scientific publications and the way they are used. Articles become known through citation for a single contribution to knowledge: either a new method or a surprising result, but never both.
The changes to scientists' reading habits due to the Internet are similar to the distinction between grazing and browsing animals. Grazers like cattle consume grass in bulk during intensive feeding bouts. Most grass is not especially nutritious and is regurgitated later as the animals sit reflectively and chew the cud. Bulk feeding and rumination means that cattle are large and ungainly creatures. By contrast, browsers like deer are much more picky in the plants they eat and select only the greenest shoots. This means that deer consume smaller quantities of food than cattle, but are constantly on the move and spend much less time at rest. Thus, the modern Internet-era scientist may be mentally nimble as the deer is physically nimble, but lacks time for cattle-like rumination.
The Internet has undoubtedly brought great benefits to us all. At the same time, the Internet make us more specialised and compartmentalised in the kinds of knowledge we access and absorb. This is a problem is an age where interdisciplinary solutions are required to solve the complex and sometimes conflicting problems of climate change, poverty, disease and biodiversity loss. In this setting, the role of informal fora for cross-disciplinary engagement becomes even more important. Here it's harder to see the Internet as a solution because the chat room can never provide the chance encounters, nor replicate the convivial cosiness, of an old-fashioned low-tech coffee room.