I 've occasionally been asked (typically by 'novice' poets with limited self-confidence) whether what they've written is actually a poem. And indeed a highly intelligent friend recently told me 'I've written a few poems, but I'm not a poet' – an assumption I feel is open to question, if only logically.

Of course everything depends on what you mean by the words poem and poet. A couple of centuries ago, form was the primary deciding factor – whether one was writing a sonnet, an elegy, an ode or whatever. Then along came Walt Whitman and his like, and both form and subject matter became much more varied, making the decision about what qualifies as a poem more difficult. A school of thought has even emerged relatively recently that considers that anything written by someone who is called (or even self-nominated as) a poet automatically qualifies as a poem – a narcissistic inversion of Flaubert's 'L'homme n'est rien, l'oeuvre tout!' to mean 'the work is nothing, the author everything'.

But such bizarre excesses aside, I think we may lay down a few (albeit fuzzy) principles to help decide what is or is not a poem, with the proviso that we're not discussing aesthetic merit – a poem's 'goodness' – which is too subjective to pin down by any rules, however fuzzy.

The fundamental principle that underpins all such decisions is that poetry (like all other literature) is a vehicle for communicating ideas. So, regardless of the poetic form employed, the first test has to be, as the late Robert Graves stressed in one of his 1960s lectures as Oxford University Professor of Poetry – 'does this make sense?' 1 It needn't be blunt as a cricket bat to pass this test, but it mustn't leave the reader wondering what it's about. Further, unless clearly intentional (as in ironic writing) factual error is a negative sign. For example from prose, in a novel about 17th century English seafaring 2 John Masefield used the expression buzz saw – an American slang term first recorded in 1886 – and it really jars on the eye. And, annoyingly, Eliot even admits to muddling the tarot in 'The Burial of the Dead' from The Wasteland. It's a patronising cop-out to suggest that 'the reader won't notice'.

Second, if a poem ostensibly subscribes to some accepted structural form, does it adhere to that form within tolerable latitude? Departures from the agreed structure of the chosen form must be clearly justified by such factors as emphasis or euphony. So for example, within an otherwise strict structure of ten syllable lines it might occasionally be advantageous to break a single line into two separate lines of six and four syllables (or the other way round) 3 in order to emphasise a word that would thereby start the second line. But to maintain conformity with the chosen form, in such a case the rhythm, for example, should proceed uninterrupted, and the intended result must obviously to the reader succeed. This test obviously doesn't apply to free verse, but more on that below.

The third test starts from stripping out all the poetic form in use – line and stanza breaks, capital letters not implied by punctuation, optional or unconventional word or letter spacing etc. – and writing out the resulting text in linear form. The question then is – does the resulting prose text lose any significant expressive power? This can only be tested reliably by an uninvolved third party and, essentially, the other way round. They must first read the plain text and then the original poem to see whether the chosen form heightens the poem's effect or not. This rule is particularly relevant to free verse, as there's always a possibility of, like Molière's Monsieur Jourdain, failing to realise one has unwittingly been speaking prose all the time.

These tests may seem rather harsh and 'analytical', and they are. But they're a tolerably effective first-level filter to eliminate the fundamentally flawed.4 Beyond that, it's increasingly subjective. Ultimately it's the reader who decides whether a text qualifies as a poem and, if so, how well it succeeds for them. It's an intensely personal judgement over which the poet has no control. So, beyond the obvious basics I've outlined, it's not really helpful to worry about whether you've written a poem or not (or indeed whether you're 'really a poet'). The trick is just to get on with it and test the results on both your intelligent friends (who may be biased by personalities) and other folk who are less likely to be. But it's worth remembering that, at least according to Graves (with whom I concur), the key qualification for being a poet is the way one conducts oneself in life,5 regardless of whether one has ever written a word.

 [1] 'Technique in Poetry' in Robert Graves, Mammon and the Black Goddess, Cassell 1965.

 [2] John Masefield, Captain Margaret, Grant Richards 1908.

 [3] I believe this is called enjambment by the 'experts', but where possible, I try to avoid the technical jargon of lit crit and 'poem analysis' simply because it can raise a barrier to understanding on the part of non-experts. Most active poets I know are not 'lit critics' or analysts, so they aren't necessarily well versed in the jargon.

 [4] however, if you'd like specific guidance on writing very bad poetry, see Susanne Bennett's highly entertaining guide.

 [5] see Lecture 2. 1965, in Robert Graves, Poetic Craft and Principle, Cassell 1967.