Larkin was right, more or less

Historians prefer not to think about coincidence. It threatens their generalising if the resemblances between events are just accidental. Simon Szreter’s remarkable and very important book argues, in effect, that coincidence has deceived the historians of family sexuality in the period 1860-1960 – and moreover that sometimes the historians connived in their own deception. The birth-rate per family in England and Wales declined ever more steeply in this hundred-year period, and it declined with roughly the same timing and speed in most other European countries. One can see how historians would dearly love the whole story to be one unitary phenomenon – which is how it is normally understood. But Simon Szreter now argues that their cherished unitary fertility decline is riddled with coincidence. The appearance of one effect linking bedrooms of 1860 with those of 1960, and English bedrooms with those in Finland and Spain, is illusory, according to Szreter. If he is right, he has completely rewritten this tract of English social history, and created a model for enquiry into the subject.

It is possible to exaggerate the degree of belief, among historians, in a unitary fertility decline. For a long time it was indeed supposed that around 1860 married couples had started doing most of the things to prevent conception which they do for that purpose nowadays: that is, they had prevented sperm ejaculated in the vagina from reaching an ovum by blocking it or killing it or washing it out. But more recently, historians have come to doubt if late 19th-century birth-control was in fact achieved by such barrier methods. They have also performed close studies of particular groups which suggest a great diversity within the courtship and bedroom practices in English society at any one time, and which point to birth control as a strategy resorted to, under certain conditions, for as far back as the historical record stretches.

As Szreter points out, such developments in the historians’ thinking should have shaken the whole theoretical edifice of a pan-European homogeneous fertility decline to its foundations. We now recognise that pre-20th-century birth-control was not a matter of modern-minded couples taking advantage of newly available rubber goods and chemicals, but we still assume that the greater part of the fertility decline (at least from around the beginning of this century down to our present) was such a phenomenon. The appeal of this story has much to do with its contemporary relevance – that is, its relevance to the world’s currently developing countries – and the power of the international agencies concerned with promoting birth-control in these countries.

The whole mistaken picture of a monolithic uptake of birth-control across modern Europe goes back to a celebrated early official attempt to demonstrate that there had been a steady and beneficial diffusion of such techniques through English society. The census of 1911 is often called the ‘fertility census’, because (as with the attempt in the 1851 census to assess patterns of religious observance) the census-forms contained special additional questions. Households had to report on how many children had so far been born into unions. The staff at the General Register Office, under the direction of Dr T.H.C. Stevenson, superintendent of statistics, worked on these data over the next decade or so. Stevenson sought to analyse the figures according to a particular categorisation of English social classes. This class-scheme was to prove momentous beyond anything he could have foreseen; at the time, its relation to fertility was simply a very pleasing confirmation (almost ironically complete, in fact) of what he already believed about the nation’s sex-life.

Stevenson’s scheme was nothing less than the five-tier, one-dimensional, occupation-based division of classes which remains, in essence, orthodox and official in the present day, eighty years later. Stevenson’s version went as follows: I Professional, II Intermediate, III Skilled Manual, IV Intermediate, V Unskilled Manual. There have been lots of complaints from historians over the years about the inadequacy of this list, but it remains ascendant. Modern officialdom has modified Class III to include ‘Skilled Un-manual’, and quite recently, as Szreter observes, Whitehall has given ‘positive identities’ to Stevenson’s ‘Intermediate’ Classes II and IV by labelling them ‘managerial and technical’ and ‘partly skilled’.

Stevenson’s emphasis on occupation and skills had at first been a response to the agenda set by eugenicists, whose hereditarian theories were being increasingly resisted, according to Szreter, by a ‘confident, revitalised and more comprehensive environmentalist analysis’ in institutions of social policy such as the GRO. The eugenicists said that low skills and high fertility were linked, leading to ‘race suicide’. There is no record of what Stevenson felt as he tested this theory, but given his environmentalist views he may have been dismayed when he saw that the linkage predicted by the eugenicists in fact held. But he also saw an alternative line of argument, which accepted the linkage but trumped the hereditarian explanation with an impeccably environmentalist one. Birth-control was the key. It was ‘diffusing’ slowly downwards from the educated and prosperous through the less educated and poorer ranks.

A less tigerish and ambitious historian than Szreter might have let the matter rest there, after his detailed and densely expounded account of these events in Whitehall around the years of the Great War. But that is not Szreter’s style. He goes on to tackle the deeper questions which the operations of the fertility census of 1911 so clearly raise, and asks head-on: ‘Was Stevenson right?’

There is clearly an irony associated with Szreter’s attempt to answer this question. If he were younger (say, roughly, a ten-year-old boy), he could look forward to being able to answer it decisively, especially with the computer-power likely to be available in the future, because in 2011 – 100 years after the fertility census – the raw data which Stevenson used, the complete household census returns for 1911, become publicly available. By that time Simon Szreter will be well into middle age. One can see that he could not wait. He must nevertheless be tantalised by the thought of what his successors will be able to achieve. (Alternatively, if Szreter had been studying Scottish sexuality rather than mainly English and Welsh, he could have used the records for the registration of marriages and births: in this country they are open only for individual enquiries.)

Szreter is confident that Stevenson was wrong, even on his own showing. The argument involved here is somewhat elaborate. Stevenson, and demographers ever since, have held that true birth-control – in the sense of full sexual relations between partners performed with the deliberate adjunct of devices and substances used to prevent conception – will most clearly show up in the statistics in the ‘stopping’ rather than ‘spacing’ of births. Large numbers of couples will be detectable as at first producing children at something like the biologically maximum rate – and then producing no more. Stevenson claimed that ‘stopping’ behaviour was discernible as ‘diffusing’ in the English social classes across time. The published data of 1911 do not permit Szreter to check this claim for couples who, through ageing or death, had finished having families by this date (the larger category), but he is able to perform the neat trick of checking it for the smaller category of younger couples who were still producing children. We can work out if this group, at least, was ‘spacing’ or ‘stopping’.

They were spacing. They do not exhibit the hallmark of birth-control required by Stevenson. There is an obvious way to rescue Stevenson at this point, in his own despite. Why can’t spacing be a token of artefactually controlled conception, just as much as stopping? After all, there remains the basic fact, which no one disputes, that total English fertility did decline after 1860. How else is this to be explained? Szreter does not rest his case only on a refutation of Stevenson, on his own terms. He agrees that spacing of births could in principle be the result of birth-control. But he has drawn a more profound observation from the published tables of the 1911 census: that low fertility achieved by spacing correlates with late marriage. Couples of child-bearing age who were conceiving infrequently were also likely to have postponed getting married.

This is probably the most important single result to emerge from Szreter’s research, and it paves the way for his own general theory of family sexuality, in the years around and after 1911, which occupies the final third of his book. It was, according to Szreter, a ‘culture of abstinence’, influential right up to Philip Larkin’s 1963 (when ‘sexual intercourse began’), which mainly drove down the fertility of England and Wales – but probably not that of other Western countries experiencing a comparable decline.

On this account, diffusionism is out of the window. No wisdom about obtaining and using certain devices and substances percolated down from the privileged to the less privileged. Moreover, the thinking that impelled couples to resort to birth-control via ‘abstinence’ would not, according to Szreter, yield a simple correlation with social rank. Couples took steps to reduce numbers of conceptions in response to the ‘perceived relative cost’ of childbearing. This pivotal concept in Szreter’s argument is developed with great subtlety to suggest how fertility-control will crop up in a much more complicated, sporadic pattern than predicted by Stevenson’s diffusionism. Hence the wonderful mosaic of Szreter’s Fig.7:1, exhibiting ‘twenty or more distinct fertility regimes’ among the working class.

The ‘perceived relative cost’ of having a child may, indeed, be about ‘cost’ – or gain – in the purest economic sense, that is, whether a child will earn or lose its parents money, but there are benefits of a more rarefied kind to set against the costs: how much prestige attaches to fatherhood, for instance. This approach to the analysis of fertility is not novel. It is a respectable theory in modern demography, and has been applied previously to the 19th-century data: by Enid Charles in the Thirties, and influentially by the Bankses to the Victorian bourgeoisie. Szreter has put the perceived-relative-costs approach to work in detail right across English society, while exploiting the full range of the powerful notion of ‘costs’.

I have invoked Larkin’s 1963, and it may appear that what Szreter has done in this book is simply to provide academic support for a familiar modern cliché about English sexual culture in the 20th century: namely, that sexual ignorance and repression were widespread until about twenty years after the outbreak of World War Two. This cliché represents a shifting forward of a supposed divide between the repressive and the non-repressive which used to be located around the beginning of this century. To some extent it coexists with this older chronological model: it is still orthodox to call repressive sexual attitudes ‘Victorian’, but it does seem that a new chronological picture is forming in the popular historical consciousness, supplanting the old Victorian/20th-century divide.

Szreter’s book is not an annexe to this new view, or at least is not conceived by him to be such. He insists that his ‘culture of abstinence’ was not driven by negative feelings about sex, such as guilt, fear or disgust (For him abstinence was merely the English way of adjusting fertility.) And he goes further than any of the popular demonising accounts of 20th-century sexuality before P.J. Proby and Christine Keeler. His book may be thought of as an ‘anti-1911’ text in three respects: ‘1911’ in the narrow sense of the fertility census is debunked on both ideological and statistical grounds; but Szreter would also like to draw a radical corollary from the fallaciousness of Stevenson. (It is a failure to face this corollary which has caused the odd persistence of the belief in an artefact-driven, homogeneous, pan-Western fertility decline, despite the quite general recognition among historians that the first forty years of this decline in England was not artefact-driven.) Szreter wants to insist that the Stevensonian account is a fantasy about past, present and future. He is convincing on how investigations which appeared to confirm Stevenson – most notably the Royal Commission on Population of 1944-9 – failed to pick up the true extent of English couples’ resort to coitus interruptus or simply to refraining from intercourse. Nothing at all happened in 1911, according to Szreter.

The failure of historians who should know better to face this thought is perhaps linked to the continuing general prestige of the ‘modern’ moment represented by the pre-Great War years – which gave rise to Cubism, Dada, Relativity and Quantum Theory, The Rite of Spring, Sons and Lovers, and so on. We’re unable to shake off the conviction that there was a revolution in sexuality, too, even as we shift the boundary between repression and non-repression forward to around 1960. We believe in the sexual misery of our parents and grandparents, but also applaud the sexual adventurousness of Bloomsbury. In Simon Szreter’s picture, both Carrington and Stevenson are the merest froth on the surface of early 20th-century English sexuality.

It is important to grasp what he means by ‘abstinence’, in this connection. He thinks of it as closely affiliated to marriage-postponement. He also thinks of it as essentially the same kind of behaviour as coitus interruptus, the latter being a compromise arrived at by couples when an intention to abstain was not found achievable: ‘couples in British society who engaged in a regime of coitus interruptus were involved in essentially the same “game” of sexual self-restraint as those practising the various forms of conscious abstinence.’ This is an unorthodox and arresting way to see the matter, but surprisingly persuasive – in the tradition of Peter Laslett’s penetrating commonsense insights.

Couples do not have to be all that abstinent in order to achieve a useful reduction in fertility. We perhaps tend to think of coital frequency as involving a threshold, as far as the chances of conception go. We assume that some sex is just as likely to produce a conception as a lot of sex. This is a fallacy. The relation between the frequency of sex and conceiving is continuous. Even if you only manage to restrict yourself to sex once a week you will still manage to postpone conception by eight months from when it could be expected if you have sex four times a week. If you can bring the rate down to once a fortnight you are buying 18 conception-free months.

In many respects Szreter’s ‘abstinence’ fits readily into his whole account. It was not an obsessive, overly stringent, or even completely binding sexual regime. It is feasible that it survived for about a century, not coming into collision either with militant sexual rigour, represented by the Social Purity movement, or with sexual emancipation, represented by Bloomsbury and its heirs. It is perhaps feasible that it was deployed to achieve reduced fertility when the perceived relative costs of children made this desirable. But there must be doubts about this crucial point of convergence in Szreter’s whole structure of argument. Will you be abstemious in sex, even in the qualified way outlined by Szreter, simply to achieve fertility-control? Does there not need also to be cultural encouragement from a climate of antagonism to sex? And if so, isn’t Szreter’s historical picture just that of popular anti-repressive triumphalism writ large?

Discussing why ‘a culture of sexual disinclination’ was not a prerequisite for his ‘culture of abstinence’, Szreter says that ‘the balance of the demographic and cultural evidence appears to point to the greater importance of deliberate, negotiated birth regulation as a positive motive, albeit one that was mediated through a culture of anti-sexuality.’ There is something fudging about the last clause. What is it to ‘negotiate’ sexual abstinence ‘through’ anti-sexuality? Either couples agreed to refrain from doing something they both liked because they didn’t want too many children, or they felt (perhaps mutually and explicitly, perhaps not) that sex was a bit repugnant. Szreter’s combination of the two does not seem workable, once you come down to practicalities. One cannot imagine a contribution from ‘anti-sexuality’ which would not exclude or at least inhibit ‘negotiation’.

Szreter’s book is at times inelegant, lapsing into repetition and clottedness (‘efforts at causal exegesis do not therefore need to be premised on the idea that some kind of absolute metastasis needs to be explained’). On the other hand, some of the writing is crisp, even witty; and this is a book of wide compass in the information it contrives to bring together. Here is a trained statistician who is also completely versed in the non-quantitative literature, and eager to bring the two into connection – and who, for good measure, incorporates a whole history of the development of the diffusionist orthodoxy in this century.