When President Barack Obama sacked General Stanley McChrystal in June because of the military man's reported indiscretions in a Rolling Stone article, no one thought much about the literary implications.

But among all the details excavated about the story that contained McChrystal's ripe comments, it is worth noticing its form as well as its content. "The Runaway General" is a prime example of narrative journalism, which has been growing in popularity and esteem.

No one can decide what name to give this kind of writing: in Europe, "reportage" is common, while in the US the term "literary journalism" has gained ground and "creative non-fiction" has some currency, especially in ÖйúAƬ. What they all share is a concern for the telling of true stories with the concrete detail of dialogue and description - the classic "show not tell" of good narrative prose.

Often, part of the definition is that the story works on more than one level, so that the specific subject matter leaves openings to other, more universal themes. And it usually tolerates - and makes plain to the reader - a greater range of uncertainty about what the writer knows and how s/he knows it.

ÖйúAƬ

The exacting demand for both detail and accuracy involves immersion reporting; at the very least, an eyewitness account of some kind. The author is writing from personal experience, albeit an experience that has been deliberately expanded and turned outward.

The Rolling Stone article is the latest - and in the greater scheme of things, not the best - example of this ambitious form of journalism, which still gets more time and space in the US than it does elsewhere. The reporter, Michael Hastings, spent a month with the general and had 8,000 words to explore his topic, which was the desperate flaws of the current Afghanistan strategy. The action kicks off in the first sentence, when the general demands: "How'd I get screwed into going to this dinner?"

ÖйúAƬ

The impact of the story is not just in the choice quotes; it also comes from an accumulation of scenes and dialogues in which the actors speak (and condemn themselves) in their own voice, giving their words and behaviour more weight. It is hard to imagine the same level of revelation in a conventional news feature. It is the form that makes the content possible.

The worry that conventional reporting practice is leaving an important gap in our understanding of the world around us, at a time when the need to make sense of it is greater than ever, has shifted in the past few years from a minority stance to a broad consensus.

The question arises: does literary journalism offer a way to fill the gap? If the selection for journalism prizes is anything to go by, the answer may be "yes".

At a conference on literary journalism held at Roehampton University in May, a study of awards for international reporting outlined by Marcel Broersma of the University of Groningen in the Netherlands and Verica Rupar of Cardiff University concluded: "it is narrative rather than discursive reporting strategies that carry the highest cultural capital in today's journalistic field".

Conference attendees also heard of a study by Miles Maguire of the University of Wisconsin, in which a memoir written by war reporter Dexter Filkins was compared with his original news reports of the same events; it turned out that the more "factual" text missed out key details, which surfaced only in the subjective form. The conclusion was that his literary journalism was actually more fact-intensive than his hard-news reporting.

Sometimes narrative journalism is offered as the answer to a crisis of legitimacy in literature as a whole. In June, The Guardian published a review by Geoff Dyer of Sebastian Junger's book War, an account of eight months spent with a US Army platoon in Afghanistan. Dyer praises the use of characterisation, observation and narrative drive in the service of non-fiction. "Reportage, long-form reporting - call it what you will - has left the novel looking superfluous," he writes, echoing a similar challenge to fiction made nearly 40 years ago by Tom Wolfe in The New Journalism.

Since I spend a large part of my days teaching the practice of long-form narrative non-fiction, I find this kind of recognition encouraging. Despite its popularity, the genre often falls below the radar. But in some ways this kind of praise only helps to perpetuate the problem.

People need to see a pattern in order to make sense of something. Literary journalism is often lost in the picture because it does not fit the prevailing paradigm. Dyer and others, such as David Shields, whose anti-fiction manifesto Reality Hunger (2010) he cites in support, claim to offer an alternative paradigm by questioning genre boundaries. Dyer characterises reportage as non-fiction prose that has "the narrative shape and moral resolution of fiction"; literary non-fiction is good because it is like fiction. Shields calls it "a blurring (to the point of invisibility) of any distinction between fiction and non-fiction: the lure and blur of the real". Literary non-fiction is good because it really is fiction.

ÖйúAƬ

But this is only true if one assumes that characterisation, detail, multiple layers of meaning and narrative drive somehow belong to the novel; or at least, to fiction. It assumes that if a writer uses craft to create a particular effect in the reader, the text loses a distinctive relationship to the real. Shields, in his book, goes further and says flatly: "Anything processed by memory is fiction." This is not blurring genres; this is the swallowing of one genre by another.

To his credit, Shields understands that in the longer history of narrative prose, these storytelling techniques first emerged in non-fiction forms; it was their adoption by realist fiction that made the novel new. But why presume that the deliberate "making" of a text equates to fictional invention? Just because we question whether language is the transparent windowpane of George Orwell's description, it does not follow that all translation of experience through language is impossible.

A more radical approach may be to argue for a reclaiming of imagination from the exclusive territory of fiction. In this new paradigm, the poetics of fact, one would understand that imagination is not about invention as such, but about the creator acting as a shaping consciousness; noticing things, imagining new possibilities about the ways in which they may be connected and then communicating that to others.

The experience of teaching in ÖйúAƬ makes this question particularly acute. One cannot teach writing practice unless one recognises that all writers of all genres make decisions - explicitly or tacitly - about how to shape a story. The aim of a course is typically to help students develop a sense of what is possible and understand the choices that arise along the way.

The provisional nature of a text is one of the most important things that a new writer can learn. One is looking at the text not as a final product, but as something in the state of becoming. Creative writing - as a discipline as well as a practice - embraces not just the work of putting words on the page, but also the reflection, preparation and research that comes before, and the waves of editing and revision that come after.

At the same time, one learns about the creativity that lies in shutting down possibilities and working within constraints. In each genre, the constraints are different; the main one of non-fiction is the demand for a process of discovery that is documented.

Dyer and Shields are right to appreciate the importance of this moment. The struggle to work out a new deal with reality is arguably the struggle of our times. Truth (veracity) is not enough; there must be authenticity as well.

Literary journalism offers one possible strategy to achieve that, in the sense that it confronts what analysts such as John Hartsock, in A History of American Literary Journalism: The Emergence of a Modern Narrative Form (2000), call the "alienated objectivity" of conventional journalism. The argument is that only by finding a way to acknowledge the subjectivity and uncertainty that exists in factual discovery is it possible for the writer to anchor this to external reality in a way that is persuasive and trustworthy. Authenticity is also on guard against what one could call alienated subjectivity; the emoting-to-order of much popular confessional or opinion journalism.

Another strategy is to privilege what Shields describes as "raw material, seemingly unprocessed, unfiltered, uncensored and unprofessional". At present, the most popular place to look for it is the blogosphere, where the web took a narrative turn. The form's weaknesses are well documented; at its best, however, the internet provides a new channel of distribution, in which old talent has the room to experiment and new talent has a new way of finding its audience. The result is a growing interest in innovative forms such as the lyric essay, and a fresh angle on voice, tone and point of view.

At the same time, the quest for authenticity in the blogosphere - expressed by the new-media pioneer Dave Winer in his 2003 Harvard lectures, when he defined the blog as the "unedited voice of a (single) person" - should alert us to a potential new idealism that is sometimes just as questionable as the much-derided ideal of objectivity.

The hunger for the "authentic" also has a dark side; an automatic distrust of anything "official" or professional that encourages a paranoid style that is both aesthetically and intellectually unsatisfying.

David Aaronovitch's Voodoo Histories: The Role of the Conspiracy Theory in Shaping Modern History (2009) documents the dangers of this form of disengagement and the ways in which it is enhanced by the peculiarities of the internet as a medium. I would add only that it is not just the technology of the medium that enables the paranoid style, but also some of the conventions that have grown around its writing culture.

Literary journalism performs a service in allowing us to acknowledge the importance of subjective feeling as well as fact; but also in recognising that while emotions are real, they do not exist in a vacuum. Authentication is not enough; we also need verification. The extreme privileging of "authentic" feeling, without the possibility of an external reality check, can lead to great violence being done to oneself and others, or at the very least, to a lack of critical engagement with those who are different in some way.

As many scholars noted at the conference in May, one must never stop asking questions about literary journalism - its creators, readers, contexts and forms. But hopefully, a growing number of people will be interested in the debates that it makes possible.

Ever since the first papyrus scroll was issued, authors have bemoaned the obtuseness of reviewers. In A Fable for Critics (1848), James Russell Lowell, the American poet, wrote: "Nature fits all her children with something to do/He who would write and can't write, can surely review."

ÖйúAƬ

To this time-honoured tradition of authorial lamentation, we may now add Herminio Martins' complaint in Times ÖйúAƬ ("Drawing the venom from the poison pen of rancorous reviews", 13 May) that the academic book review has degenerated into a "blood sport", an outlet for the "glib, vacuous, prejudiced, obtuse, frivolous, ignorant, petty-minded, rancorous or ill-mannered".

He offers a number of solutions. The first is to institute a code enjoining reviewers to "report the book's content and avoid a disparaging tone". He also urges reviewers to provide "contact addresses to enable correspondence with reviewed authors". But his most draconian option is to "dispense with book reviews in academic journals altogether".

The last is a startling proposition. Is the academic book review really in so parlous a state as to warrant extermination? Surely a representative sampling would for every tendentious review yield many more displaying professionalism, proportionality, fairness, rigour and insight. We should not, however, gainsay Martins' laudable desire to improve the level of scholarly book reviewing. The question is whether his proposals aid us much in achieving that end.

To my eye, at least, Martins' solutions oscillate between excess and insufficiency. To abolish the academic book review tout court so as to eliminate bad ones strikes me as analogous to abolishing medical operations so as to reduce malpractice lawsuits, or putting an end to democracy so as to avoid electoral corruption. This is to throw out the baby with the bathwater.

As for a code of book reviewing, we might do well to have one, but if it proscribes disparagement, it may well have the unwelcome effect of preventing a reviewer from faulting a meretricious book. One of the primary functions of reviewing, after all, is to sort wheat from chaff, which requires permitting chaff to be called what it is.

To publish reviewers' contact information with reviews, lastly, is anodyne and unobjectionable, but surely bolder experimentation can be imagined. What if, for example, journal publishers were to follow THE's lead by allowing rejoinders by authors, posting them after the book review on their websites, along with comments from readers?

To be fair, Martins proposes that reviews be replaced, not merely abolished. What he favours are major symposia involving multiple commentators on certain works, along with brief book notes conveying the tables of contents of other notable books. The problem is that few books rise to a level meriting a symposium, and half-paragraph summaries cannot do justice to the remainder. Few journal editors, therefore, will rush to adopt Martins' recommendation to eliminate book reviews completely.

How may we better meet the legitimate aspiration for higher-quality book reviewing?

Postgraduate education may offer a viable instrumentality for instilling reflection on the craft and ethics of the review. What if reviewing were more fully integrated as a subject of postgraduate studies in the humanities and social sciences?

In core modules, instructors could assign a significant scholarly book, requiring students to locate five to 10 reviews of it in academic journals and write an evaluative essay about which reviews seem best, which worst, and why. Alternatively, students could write their own reviews of the same book (hewing to the standard 500- or 750-word limit), then circulate their reviews and discuss them. A third approach would be for the whole seminar group to read and discuss all the reviews in a given issue of a leading journal.

At the school, departmental or college level, workshops on book reviewing could be convened. Panels of faculty from multiple disciplines could discuss what they believe makes an exemplary review and what pitfalls to avoid in writing them. Such exercises would reveal that there is no uniform format for a successful book review. At the same time, they would help embryonic scholars achieve a greater self-awareness as future book reviewers and provide them with pointers for success.

What guidance should we transmit to students in such contexts? Here are 10 points that I believe make for an excellent scholarly book review:

- It is based on a close, careful reading of the book. (This may sound obvious. It is not.)

- It accurately and succinctly conveys the book's topic, meaning, evidence, methods and positions. What is the book about? What are its key concepts? What are the data it marshals? What research does it reflect? What is its scope? What are its central themes? What is its thesis?

- It provides a tactile sense of the author's style by supplying apposite quotations from the text, not relying on paraphrase alone.

- It contextualises the work. Who is the author and how does this book relate to his or her oeuvre? Does it reflect a distinct theoretical or interpretative tradition or school? Do other books exist on the same topic? How does this book's approach contrast or compare with them?

- It identifies the work's significance. What is original or unique in it? Does it contain new evidence? Does it make us reconsider well-traversed ground by offering a new interpretation? Does it portend new directions in the field to which other scholars should be alert? Does it relate to wider cultural or political developments in society as a whole? Does it alter the way we view its subject?

- An excellent review will be evaluative. It will make critical judgements about the quality of the book's research, the logical coherence of its argument, the agility of its presentation and the extent of its intellectual contributions. The review should take a sounding, that is, not merely list the table of contents; in fact, chapter-by-chapter synopsis is the dullest imaginable form. Instead, the review should evaluate the book's merits or weaknesses. Are the illustrations, tables, charts and footnotes in order? Does the text deliver on the title? Is the methodology valid? Is the work outstanding or disappointing, or both in equal measure?

- It aspires to audience-appropriate clarity. Readers of specialised journals such as the Kantian Review or Ancient Mesoamerica may not need to have explained some things that may strike others as arcana or jargon. Nevertheless, the best reviewers in every field strive for translucent expression, presuming in their mind's eye a highly educated generalist reader, while also bowing to the interests and needs of the journal's core readership.

- It is fair and measured, not tendentious. Judgements are supported by examples. Criticism is not niggling but substantive. Authorial insight is flagged, as are flashes of brilliance.

Typically this implies a temperate if not courteous manner, but at times the responsibility of the reviewer to uphold intellectual standards makes ineluctable a polemical or disdainful expression. Honest, vigorous dissent from a thesis can be an invaluable service. Readers may be trusted to discern that the reviewer is giving expression to a particular viewpoint. The world of ideas is at its best when it achieves what Diana Trilling - a critic of scathing ferocity when she edited The Nation magazine's book review section in the 1940s - once called "a life of significant contention".

- It is written with verve and distinction, its words well chosen and compelling, its transitions affording a smooth flow of ideas. At its very best, "academic" means learnedness, not pedantry - although if that ideal were always realised, the likes of Kingsley Amis and David Lodge would not have had nearly so much fun at our expense. Regular reading of this publication, the Times Literary Supplement, the weekend Guardian, The Observer and other literary and cultural publications can help to enhance book-reviewing sensibilities - not by literal appropriation, but by the internalisation of standards and apprehension of modalities.

- It often transcends its particular book and subject by using them as springboards to vistas beyond. Scott McLemee, who writes the Intellectual Affairs column for the American website Inside Higher Ed, is a particularly adroit writer in this manner.

All of this is to say that the scholarly book review should be cultivated, not discarded impatiently. Reviews, even the little-read ones in scholarly journals, are critical to the vitality of a fragile book culture threatened by technological competitors and consolidation in global publishing. Review-writing enhances analytical skills relevant to other academic tasks, such as peer evaluation of teaching, refereeing of journal articles and external oversight. Finally, book reviews are arguably the most important way in which scholarly publications are measured by the wider profession.

Universities can and must find more concerted ways to instil in future scholars an awareness of the ethics, varieties and potentialities of scholarly book reviewing so as to ensure that academic reviews are written at an optimal level of craft, art and responsibility.

ÖйúAƬ

Christopher Phelps is associate professor in American studies, University of Nottingham.