Exclusive: Ted Hughes’s poem on the night Sylvia Plath died

The New Statesman publishes a previously unseen work by the late poet laureate.

The New Statesman publishes a previously unseen work by the late poet laureate.

In tomorrow's New Statesman, which has been guest-edited by Melvyn Bragg, we publish a previously unseen poem by Ted Hughes. "Last letter" is a poem that describes what happened during the three days leading up to the suicide of his first wife, the poet Sylvia Plath. Its first line is: "What happened that night? Your final night." -- and the poem ends with the moment Hughes is informed of his wife's death.

Hughes's best-known work is 1998's Birthday Letters, a collection of poems that detail his relationship with Plath. Though the published poems make reference to Plath's suicide, which occurred in February 1963, when she and Hughes were separated but still married, none of them addresses directly the circumstances of her death. This, then, would appear to be the "missing link" in the sequence.

The earliest draft of "Last letter" held in the British Library's Ted Hughes archive appears in a blue school-style exercise book, which is believed to date from the 1970s. The book contains drafts of several poems that appear in Birthday Letters. A more refined draft of the poem is found in a hardback notebook. After drafting poems by hand several times, Hughes would usually type out poems when they were near completion, adding notes in the margin where necessary.

Below are images from various drafts of the poem:

Add. 88918/1/6, f.1

The image above is of the first page of the earliest known draft of the poem, which went through many revisions before the final version appeared

2010+40ted poem 2

The image above is the first page of a later draft of the poem (date unknown)

Add. 88918/1/8, f.11

This image is from a draft of the poem contained in a hardback notebook. As is evident, Hughes would extensively rework phrases and add lines throughout the various stages of drafting. When a poem was finished, he would usually type it out, annotating with comments where necessary

In a letter from 1998 to his fellow poet Seamus Heaney, Hughes says that he first started to write simple verse "letters" to Sylvia Plath in the early 1970s. Hughes began writing them piecemeal; later he tried to do it in a more concerted way but found that he couldn't, so he went back to writing them occasionally. Some of the Birthday Letters poems appear in the 1995 New Selected Poems, but in correspondence with friends (also held by the British Library), he says he had found some of the other poems too personal to publish at that time.

Tonight Channel 4 News covered the story and recruited the actor Jonathan Pryce to read a section from the poem.

To read the poem in full, pick up a copy of Thursday's magazine.

Daniel Trilling is the Editor of New Humanist magazine. He was formerly an Assistant Editor at the New Statesman.

BBC promotional material
Show Hide image

Predicted the death of broadcast TV? Then have I got news for you

Death by a thousand Netflix clicks? Broadcast viewing figures tell a different story. 

Broadcast TV is dead. Television ratings have been declining since 1981 at the latest, and have more recently entered a terminal spiral. 1981 was when the creation of the Broadcasters Audience Research Board (Barb) established a single method for calculating how many people were watching any television channel. This made the number of viewers of BBC and ITV programmes directly comparable for the first time.

Barb endures, its methods updated, and it’s now used for all channels from BBC One to QVC HD +1 and everything in between. It’s because of Barb that we know that the most-watched programme of 2015 - The Great British Bake Off Final - managed 16.03m viewers, and that this is just over two thirds of the audience of 23.3m that saw the most-watched UK TV transmission of 1981. (It wasn’t what you think it was, by the way, we’ll come back to that at the end.)

Yet, despite the explosion of multi-channel television in the thirty five years between those figures, the most-watched television programmes of any year are now, as then, on BBC One and ITV. In any given week, only BBC Two and, even less frequently, Channel 4 might have a chance of pushing into the weekly top ten, but anything further down the programme planner will be even further down any integrated chart.

“Ah, but,” people say, “It’s all about streaming, iPlayer and online now, isn’t it?” Well, those things certainly exist, and because of them Barb’s Project Dovetail, launched in 2016, incorporates data from broadcasters’ online operations into their standard reports. Dovetail, though, only applies to online arms of recognised broadcasters, including those that didn’t exist when Barb was created. Sky’s channels are assessed through both this and more traditional methods. It’s from them we know that fewer people in the UK watched Sky Atlantic’s most-ever-watched episode of Game of Thrones than did that same week’s episode of The Great British Sewing Bee on BBC Two. Including time-shifting, SkyGo and in-the-week repeats.

However, it’s undeniably true that we don’t know how many people are watching Netflix or Amazon Prime’s original material. Statements are occasionally made, but data isn’t shared. There are no agreed points of comparison. We do know that Netflix has c5m subscribers in the UK. So, hypothetically, if three people per account watched the same programme, it could add up to the viewership of the Bake Off Final. Given that BBC One’s numbers were drawn from a potential viewership many times the size (because tens of millions more people have tellies than have Netflix) it doesn’t seem likely. Frankly, if it’d happened they’d be hollering about it from the rooftops.

Intriguingly, the US company Symphony AM, which uses a technique not dissimilar to Barb’s to calculate US viewership, once produced its own figures for US (not UK) Netflix use. They calculated that there were around 8.5m US viewers watching Netflix’s biggest series, Fuller House, a sequel to a 1990s US sitcom with little recognition factor in the UK. Netflix called this "remarkably inaccurate", but declined to produce any figures of their own in rebuttal.

Meanwhile, Barb’s first ever Project Dovetail report concluded that Cuckoo, Happy Valley and Call the Midwife were the most in-demand of "on-demand" programmes, all three being BBC series that were already doing well. There aren’t, it seems, programmes that no one watches on transmission, and which gain most their viewers via catch up. (The arguable exception is imports that simulcast with US transmissions, being nominally shown at 2am before being "repeated" 20 hours later.)

Timeshift changes the numbers, but not what we watch. Something doesn’t sneak up from behind the back because all its viewers, perhaps from a different (assumed to be younger?) demographic, are watching by a different method. An example: Sherlock’s 2014 premiere created an all-time record for timeshifting: 3.5m viewers. That’s almost a third of its audience, arguably a significant chunk, but it’s also inherently much less than half. Most people who saw the programme saw it when it was broadcast. Not afterwards.

If you ignore 2012, where the London Olympics’ Opening and Closing Ceremonies pulled in extraordinary numbers of viewers, the most-watched television programmes of the year have seen a small, but definite, upswing in their ratings this decade, starting at 13.5m and building to 16.03m. This suggests either that television viewership is actually increasing, or that viewers who decamped to other methods of watching are being included again thanks to methodological changes. The repeated, stunning ratings success of the New Year’s fireworks on BBC One may point to the former being true, larger audiences being a consequence of the reduced entertainment budgets of many in austerity Britain.

So, the programmes that are most seen by shiny new methods are the same ones seen on transmission, and those methods deliver fewer viewers than dusty old-fashioned broadcast? Still. Why is that? Live events, such as football matches, finals of talent shows and Royal occasions have always featured strongly in weekly and yearly top tens of programmes. Such things are generally watched live or not at all. A sense of communal watching, of shared experience, is only possible by everyone doing it at the same time, and while that inevitably means on broadcast, it no longer has to involve everyone socialising being in the same room.

Social media platforms, accused so often of driving people apart, merely provide another means of human contact. Certain corners of Twitter and Facebook come alive during televised football matches, especially those on BBC and ITV. And such things aren’t confined to live events. BBC Four’s Top of the Pops repeats routinely cause #TOTP to top its list of UK trending hashtags as viewers try and top each others’ jokes about acts’ hair or clothes, or indulge in nostalgic reminiscences of the era the episode is from.

Perhaps surprisingly, this online communalism does extend to drama series, despite the obvious fact that if you’re writing and tweeting jokes about a television drama series while watching it, you’re not paying enough attention to it. It may be because of things like this that this year already four drama series on BBC One or ITV, Broadchurch, Call The Midwife, Sherlock and the Moorside have been seen by more than 10m people, and several more have come close to it. In early January, when the New Year’s Day episode of Sherlock reported overnights of 11.33m I confidently asserted it would be the most watched drama series of 2017, as it had been in 2016 and 2014 and 2012, but now I’m not so sure.

The UK is the historical home of television. The pre-war BBC Television Service is usually regarded as the first regular television channel and we watch, proportionately and in absolute terms, a lot of it. The 10m viewers enjoying the BBC's biggest dramas represent more than a fifth of British adults. Compare that with the USA, say, where the two biggest series on television (inexplicably The Big Bang Theory and NCIS) draw around 19m viewers a week from a population five times the size of the UK’s. Television is very much our thing, and it seems like we’re not letting it go, however much those who try to predict the medium’s future want us to.

Online hasn’t killed terrestrial broadcast TV, anymore than multichannel, catch-up or VHS did. And in each case, there were plenty who were certain it would. New delivery methods account for far less of our viewing habits than terrestrial television does, and by far the largest share of terrestrial television is still seen on broadcast. Short of another London Olympics, we’ll never again see numbers like those gained by the 1981 premiere of Jaws, which knocked that year’s Royal Wedding off the top of the yearly chart, but broadcast TV is not dead.

It isn’t even resting.