Rumours of a death that may be true

In the humanities and social sciences, the monograph is the standard means of moving forward. It not only communicates the results of original research, it also presents a body of new knowledge to a wider public, both within a specific discipline and beyond it. It offers a sustained argument, substantiated by primary evidence and pursued in a carefully articulated and extended discussion in which the new material is allowed to speak for itself.

Monographs are different from textbooks, from syntheses for a more general public and most particularly from the common hybrid ever more frequently favoured at the monograph's expense - the book based on good ideas but resting on meagre empirical research and thrown together as a result of commercial and research assessment exercise pressures. Articles, moreover, cannot act as substitutes for monographs so much as complement them. They can and should communicate new research, but are particularly suited to offering preliminary studies and exploratory arguments, to introducing an aspect of a larger study in the process of development, to announcing a new piece of evidence that throws new light on a well-known issue and to the analysis and revision of scholarly debates. One undergraduate said: "I like articles because they are written by people who've read the monographs."

As an editor of the Cambridge Studies in Medieval Life and Thought, the Cambridge Studies in Palaeography and Codicology and the Utrecht Studies in Medieval Literacy, I assess many excellent book proposals and book-length typescripts. I also assess many non-series monographs for Cambridge University Press and other publishers in Europe and North America.

On this evidence, while the monograph is flourishing, it is being threatened from two directions. The first is the insistence by some university departments that its members concentrate on producing articles, presumably in order to make a greater contribution to the arithmetical calculation of the RAE, in spite of the RAE's recent claims to be assessing quality rather than quantity. Second, publishers are clearly in a difficult position as reduced library budgets mean the number of monographs bought is declining, print runs are cut back, monograph prices rise and they become less attractive to buyers, including academics. Some libraries have come up with the bright idea of having books on a specific subject taken by one within a group of libraries, forgetting that it is the scholar who will then subsidise those libraries with his or her time and the expense of travelling to the library that has the book.

Perhaps monographs are being published too expensively, or are not being marketed in the most sensible way. Do different subjects respond to different kinds of marketing? It is interesting that in tightly organised subject areas such as classics, Byzantine studies and medieval history, sales are buoyant, whereas in more disparately organised subjects such as modern British history, sales have slumped.

It is certainly becoming more of a battle to get primary scholarship published when commercial gain is regarded as more important than contribution to knowledge. Too many marketing departments of academic publishers are driving what ought to be editorial and academic decisions and thus obscuring the distinctiveness of academic presses. Promotion of pioneering research is incompatible with reactive marketing policies, which predict sales of a book on the basis of a superficial assessment of sales of earlier books perceived to be on related or similar subjects. The marketing principle appears to be that more of the same is safe, but anything new is too risky. Editors must continue to trust their scholarly instincts, recognise what is new and promote its publication in partnership with the academic publishers. Good university teaching and good research require constant injection of new ideas and new knowledge.

Rosamond McKitterick is professor of medieval history, University of Cambridge.