Stephen Krashen, Language and Language Teaching 1,2: 38-39. 2012
When we ask the time, we don't want to know how watches are constructed. Georg Christoph Lichtenberg (1742-1799)
Our current journals in language education are full of long papers. A typical journal might have, at most, five major papers. Sometimes we have to write long papers, but much of the time, it's unnecessary: the papers often contain long introductions more suitable for doctoral dissertations or review ("state of the art") papers, apparently designed to provide evidence that the author is well-read, and long conclusions, with a repetition of the findings and the author's detailed and lengthy speculations about what the results might mean for theory and application.
Readers of professional journals don’t need this. Introductions should only give enough to alert the reader to what the article is above, and provide a few citations in case the reader needs more information. If the articles cited in the introduction are readily available, readers are free to consult them, and a brief indication of implications is generally more than enough for experienced readers. Also, if the results section is clear, no repetition of the findings is necessary in the conclusion.
Watson and Crick's Nobel Prize winning paper on the double helix was only one page. Their conclusion: "It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material" (p. 737).
Long papers drain intellectual energy from both readers and writers and waste their time.
They take longer to write, and much of the energy in writing them is dedicated to sections that don't engage the writer: Writing is a powerful tool to solve problems, and can result in substantial cognitive development (it can make you smarter), but to do this, the writing must be directed at a difficult problem (Langer and Applebee, 1987).
Long papers take longer to read. Even readers who try to skim long papers have to devote time and energy to find the essential parts, and run the danger of missing the details.
A Disservice to the Profession and to the Scholar
Too-long papers hurt the spread of knowledge in two ways: They waste our time in both reading and writing, and they promote sloppy reading. Many readers are content just to read the abstract and perhaps the summary of technical papers, with a glance at a table. This means, of course, that significant details on methodology and the analysis, buried in the paper, are missed, and crucial points and often errors are perpetuated.
Too-long papers also take up space. A journal with five too-long papers could easily include 20 short papers. This space limitation hurts the dissemination of knowledge, because less genuine information is available, and makes it much harder for junior scholars to publish and to get tenure and promotion, especially when universities require publication in certain journals. This problem will be alleviated as more journals are done on the internet, of course.
Again, sometimes papers have to be long. But often they don't, and the problem is usually long introductions and conclusions that go far beyond the needs of the paper.
Language education has clearly taken its tradition from the humanities, which favors dissertation-style prose, rather than the sciences, where papers are usually much shorter.
It is probably no coincidence that citation rates in the sciences are much higher: Hamilton (1991) reported that about 91% of papers published in atomic, molecular and chemical physics, and 86% in virology had been cited at least once. In language and linguistics, only 20% had been cited and in American literature, less than 1%.
Watson, J. and Crick, F. 1953. A Structure for Deoxyribose Nucleic Acid, Nature 171, 737-738.
Hamilton, D. 1991. Research papers: Who's united now? Science 251: 25.
Langer, J. and Applebee, A. 1987. How Writing Shapes Thinking. Urbana, IL: National Council of Teachers of English.