One of the many paradoxes of higher education is that while it is dedicated to the production of new, often ground-breaking ideas, its institutions perpetuate obsolete practices from previous centuries. Nowhere is this more evident than in academic writing.

Think of a typical journal paper in the humanities or social sciences. Most of its content is part of some ongoing intradisciplinary dialogue, responding to and building on others¡¯ work. And this is de rigueur. If you submit a paper without footnotes, this will not be taken to be a sign of refreshing originality but an indication that you have not ¡°engaged with the existing literature¡±. Rejection is inevitable.

That may be perfectly defensible. But the way writers engage with each other is deeply odd. Time and again, papers identify a lack of clarity, an ambiguity or a contradiction in the work of others. They then go on to try to clear up the alleged mess. But if something really is puzzling in the work of a colleague, why not just ask them about it first? Instead, academics write as though they were still living in the 17th or 18th century Respublica literaria, the Republic of Letters, in which hand-written correspondence was the only way to exchange ideas.

One of the republic¡¯s central institutions was the informal Acad¨¦mie Parisienne, established in 1635 by the polymath monk Marin Mersenne to circulate copies of works transcribed by his own hand to a network of about 140 intellectuals via horse-borne couriers. Among the fruits of this system were the objections and replies to Descartes¡¯ Meditations. The French polymath Pierre Gassendi, for example, objects: ¡°You deny that a dog has a mind like yours, but it certainly makes a similar kind of judgment when it sees not its master but simply the hat or clothes that he is wearing.¡± Descartes replied, no doubt several weeks later: ¡°I observe no mind at all in the dog, so I don¡¯t think there is anything to be found in a dog that resembles the things I recognise in a mind.¡± Change the second-person form of address to the third person and you have the exact same drawn-out form in which modern disputes are still conducted.

É«ºÐÖ±²¥

I¡¯ve learned to follow these conventions when writing for journals or collections. But when writing for a wider audience, I wouldn¡¯t dream of discussing someone¡¯s work at length without speaking to them first. I read up on what people say and then interview them, putting my questions, objections and clarification requests to them directly.

This has several advantages. First, it minimises the risk of getting the author wrong. Second, it gives them the opportunity to clear up anything that might be unclear or widely misunderstood. So my piece can not only set out an objection but accurately describe how the thinker responds. Third, if I have any thoughts about how to take these ideas further, I can test them out on the person best placed to judge whether my approach is fruitful. All this increases the chances that what I write gets to the heart of the matter.

É«ºÐÖ±²¥

It seems to me obvious that academic papers would benefit from the same approach. Instead, people continue to publish according to a system whereby papers generate papers and it can take several to get to where you could have got to the first time, if you ever get there at all.

The cynic might conclude that this is the very point: academic work is measured both by quality and volume. But given that most academics despise the publish or be damned culture, this is all the more reason to reform the system.

Of course, it is not possible to discuss every question and objection. For the most cited academics in particular, the pre-interview approach would put too many demands on their time. But you don¡¯t need to discuss every single point with every single person you cite; interviews need to be reserved for when they are most needed. For the potential interviewee, meanwhile, the time it takes to talk has to be balanced against the time it would take to write responses that set the record straight, as well as the irritation.

Another objection might be that our current conventions provide a clear and transparent document trail of what was said, while no one could check what was said in a private interview. This may sound serious but if the right protocols are followed, it just doesn¡¯t matter. As long as the interviewer gives the interviewee the right to check any quotes used, what is said in the resulting paper is as fully and properly on the record as any other published words.

É«ºÐÖ±²¥

Our current system made sense in the days before cheap telephone calls and the internet. But there is no good reason for the convention to persist today. Whenever an academic writes something like ¡°It is not clear what x means by¡­¡± referees should send the paper back with the demand: ¡°So why don¡¯t you ask them?¡±

Julian Baggini is academic director of the Royal Institute of Philosophy. His latest book is The Godless Gospel (Granta).