In 1939 Oxford University Press published Lark Rise, Flora Thompson's collection of fifteen essays about a rustic childhood that had unfolded during the eighteen-eighties, in a hamlet in the ‘wheat-growing north-east corner of Oxfordshire’.  That was followed, in 1941, by Over to Candleford, which described more about life in Lark Rise and the nearby country town of Candleford. In 1943 a final volume of essays, Candleford Green, chronicled the protagonist's departure from Lark Rise, and her relocation to the village of Candleford Green, where she took a job at the post office. The three volumes were later combined to form Lark Rise to Candleford.  Since Thompson's death in 1947, that text has come to be considered as a – indeed, perhaps the – classic account of life in the late Victorian countryside. That much, at least, is suggested by its popularity, which exceeds that of the other esteemed works of rural history and literature, such as Joseph Ashby of Tysoe, The Wheelwright's Shop, Kilvert's Diary, and even Laurie Lee's Cider with Rosie.
The story of Lark Rise to Candleford is told in the third person, and takes place in a hamlet, town and village that, in name at least, are fictional. Nevertheless, the text is sometimes cited as if it were an uncomplicated work of autobiography. It is included, for example, in Burnett, Vincent and Mayall's critical bibliography of modern British working-class autobiographical writing.  The young protagonist, Laura, is usually identified as Thompson herself, the hamlet of Lark Rise is taken to be Juniper Hill in north-east Oxfordshire, Candleford the town of Buckingham, and Candleford Green the nearby village of Fringford. 
It is easy to understand why some scholars have concluded that Lark Rise to Candleford is a ‘lightly disguised’ autobiography.  The plot mirrors the essential details of Thompson's childhood in Juniper Hill, and her subsequent employment at Fringford post office. The sense of authenticity is strengthened further still by the inclusion of numerous details about ordinary country life. The first volume, for example, describes local wages, housing, dress, reading habits, food, religious sentiments, neighbourliness, travel, festivities and much more. The fact that Oxford University Press, which ostensibly did not deal in fiction, published Lark Rise to Candleford, doubtless served only to enhance the impression that it was ‘all remembered from a child's experience, all faithfully set down, all true’. 
Thompson actually cautioned readers not to regard her writings as strictly autobiographical. In the preface to the first edition of Lark Rise, she instead presented her ‘story’ of ‘life in an Oxfordshire hamlet during the ’eighties’. Years later she acknowledged that she ‘wrote more freely’ in Over to Candleford and Candleford Green, fashioning the text's characters and habitations out of her recollections of numerous disparate towns, villages and people from the region where she had grown up. It comes as little surprise, then, that ‘some sections of Lark Rise can be shown to be true, some doubtful or suspect, others untrue’.  More revealing than discrepancies about the history of enclosure in the hamlet, or the age of its inhabitants, however, are the omissions from Thompson's description of rural life. Violence, emigration and, most importantly, death receive little attention in Lark Rise, despite their ubiquity in Juniper Hill during the eighteen-eighties. (Four of Thompson's nine siblings died before the age of three, contrary to the local saying she quotes that ‘Nobody ever dies at Lark Rise’.) In Barbara English's view, Thompson deliberately ignored those rather gloomy subjects in order to idealize the rural past.  In a recent study of labouring life in the Victorian countryside, Barry Reay agrees that Lark Rise to Candleford is a ‘highly nostalgic account’. 
The wide panorama of Lark Rise to Candleford has obliged scholars to be selective in their assessment of its reliability as a historical source. The depiction of the plebeian belief in witchcraft, for example, is yet to be subject to detailed criticism. That neglect is unfortunate, because the text is highly relevant to recent historiographical developments in that area. The legal prosecution of supposed witches came to an end in Europe in the later seventeenth and eighteenth centuries.  During the ensuing decades, the crime of witchcraft was removed from the statute books, and replaced with legislation that prohibited ‘pretenses to such arts and practices’ as witchcraft.  However, the scepticism of lawmakers about the reality of sorcery was not shared by many of their subjects. On the contrary, among the population at large there remained a fairly widespread belief that certain individuals possessed supernatural powers that allowed them to afflict their neighbours with all manner of misfortunes. 
Several historians have cited Lark Rise to Candleford as evidence that the plebeian belief in witchcraft was either in retreat, or was extinguished, in England during the nineteenth century. The widespread acknowledgement that Thompson is not an entirely reliable guide to the Victorian rustic scene appears to have made little impression here, as those historians have quoted from her book as if it were an ordinary autobiography, without really considering the potential for distortion and omission. By way of redress, this article offers a critical appraisal of the portrayal of plebeian witchcraft belief in Lark Rise to Candleford. To do so it reconstructs the witchcraft beliefs and practices that prevailed in rural Oxfordshire and south Warwickshire during the nineteenth and early twentieth centuries. This was a relatively impoverished region, where agricultural wages were among the lowest in the country, Anglicanism was strong, and chronic want persisted in the villages until the early decades of the twentieth century. 
Sources consulted for this study include folklore reports, diaries, local newspapers, religious literature, gazetteers, and genuine autobiographies and recollections. Their interpretation is aided by insights gleaned from anthropological studies of witchcraft from across the globe. Since the revered works of Alan Macfarlane and Keith Thomas, some scholars have voiced misgivings about the use of anthropology to historical enquiries.  Such a collusive practice between the disciplines, its critics allege, can lead to a failure to appreciate the uniqueness of the past.  Of course, historians must be sensitive to cultural peculiarities; for example, that witchcraft in the early modern Isle of Man revolved to an unusual extent around the idea that earth and dust were carriers of magical power.  Having said that, it is also essential not to overlook the existence of transcultural, if not universal, patterns of belief. A number of recent investigations have stressed the many similarities between witchcraft beliefs over both time and space.  The Victorian folklorist George Gomme was, it seems, ahead of his time when he observed of witchcraft that ‘savage and civilized folklore are perhaps more susceptible to comparison in this department than any other’.  It is because of the existence of widespread patterns of belief, which transcend local and even national communities, that knowledge of anthropological studies can ensure one correctly interprets the historical record. Consequently this study takes the approach suggested by John Demos: that the best way to study historical witchcraft beliefs is through both ‘parochial reconstruction’ and transcultural comparison. 
There is an obvious reason to doubt the accuracy of direct speech quoted in autobiographical literature. ‘No one’, it is said, ‘can remember the exact words used in conversations that took place five, ten or twenty years earlier’.  An author who relies upon the record of a diary or journal may perhaps be trusted; but Flora Thompson, like most of those from the working-class autobiographical tradition, wrote from memory rather than notes.  Despite that, Lark Rise to Candleford confounds the conventional wisdom about dialogue in autobiographical writing – at least, as far as its depiction of belief in witchcraft is concerned. In fact, one of the few details corroborated by the historical record is a piece of direct speech. In the second chapter of Over to Candleford, the young protagonist asks her mother ‘are there any witches now?’, and is told in reply: ‘No. They seem to have all died out. There haven't been any in my time; but when I was your age there were plenty of old people alive who had known or even been ill-wished by one’.  This is not, it must be conceded, to suggest that Thompson recounts, word for word, a specific real-life dialogue. Rather her character voices something of an adage – almost a cliché – about witchcraft that actually was repeated by many people during the Victorian and Edwardian period.
We know as much because from the mid nineteenth century thousands of labouring-class people were questioned about witchcraft by inquisitive, and often socially elite, amateur folklorists.  Like the character in Thompson's text, interviewees sometimes remarked that practitioners of black magic had died out.  Others implied the disappearance of witchcraft by speaking about it in the past tense. As a result scholars like the Revd. C. E. Prior, a folklorist of the north Oxfordshire countryside, concluded that they had just missed the era when: ‘Witches were quite common’.  On his sojourns around the region in the early twentieth century, all the Revd. Prior managed to elicit from his rustic informants were stories about local witches who had died several decades ago. His experiences, in that respect, were typical. The earliest record of that convention is Alfred Beesley's History of Banbury (1841), which repeated some local ‘tales’ about witches who had danced midnight revels in the ‘olden times’.  Conducting research some fifty years later, during the eighteenth-nineties, a historian from the west Oxfordshire market town of Witney was told about witches who had communed with the ‘Evil One’ and terrorized the townsfolk during the period of the Napoleonic Wars.  Around the same time a folklorist engaged in a countywide study of ‘superstitions’ succeeded only in gathering narratives about witches who had lived ‘forty years ago’, ‘fifty years ago’, or even longer ago in the past, during the childhoods of his aged informants.  Folklorists active c.1910–1930 were also, ‘after much persuasion’, able to elicit witchcraft anecdotes from the people of the region, which were said to have occurred as late as the eighteen-eighties.  Even during the nineteen-fifties and sixties members of the Oxfordshire and District Folklore Society were still uncovering stories about witches who transformed themselves into animals, used the occult to predict and cause misfortunes, and were visited by the devil on their death beds.  And, of course, the witches in question were said to have belonged to the previous generation, which had departed this world during the first decades of the twentieth century. Similarly when, in the late nineteen-sixties, the author Donald McCormick probed some of the natives of rural Warwickshire for information about local witchcraft he was told that it ‘was either driven underground or it temporarily disappeared’ during the ‘early part of the present century’.  The dialogue in Lark Rise to Candleford was thus not wholly fictitious. When asked about witchcraft many people really did respond by saying that it had died out a generation ago.  What is so intriguing about that response, however, is that it was so persistently repeated throughout the period c.1840–1960.
Far from being unique to Oxfordshire and south Warwickshire, similar claims about the recent demise of witchcraft were made elsewhere. So much so that one folklorist of mid nineteenth-century Catterick, in Yorkshire, had obviously come to expect her informants to say something of that order, and posed her question accordingly: ‘you surely don't believe in witches now-a-days?’ Perhaps only one response was to be expected under those circumstances: ‘No! I don't say 'at I do; but certainly i’ former times there was wizards … and them sort o’ things’.  Yet even scrupulous folklorists, who were careful not to put words into their informants’ mouths, were told similar things. Rustics from Berkshire's Kennet valley, it was noted in 1902, ‘generally say “there are no such things [as witches] nowadays”’.  During the eighteen-nineties Lady Rosalind Northcote of Devon was likewise told by her elderly confidants that witchcraft ‘used formerly to flourish’ in the region.  In Lincolnshire it was the same: ‘By the end of the nineteenth century witchcraft was said to be declining’. 
In Over to Candleford the narrator refrains from directly commenting on the statement that witches ‘seem to have all died out’, and instead draws attention to the apparently common conviction that ‘Anything the Bible said must be true’.  Nevertheless, Thompson appears to have intended that her readers take the quotation at face value – as an expression of the character's sincere belief that witches no longer existed. Candleford Green, the third volume of the trilogy, explicitly states that plebeian culture in Oxfordshire had become secularized by the close of the nineteenth century. According to the narrator's gloss, the ‘old dark superstitions had gone’ from village life, and old women ‘were no longer suspected of witchcraft’.  Although some rustics remembered stories about witchcraft from their childhood: ‘In the eighteen-nineties in that part of the country ordinary people either disbelieved altogether in witchcraft, or thought it one of the old unhappy things of the past’.
In a recent study Owen Davies has concluded that such testimony ‘undoubtedly reflect[s] the decline of the witch in rural communities’ during the period.  According to his argument, the commonplace assertion that witchcraft was a thing of the past indicated the ‘undoubted general popular perception that there were fewer witches than there used to be’.  That is not to say that people ceased entirely to believe in the abstract reality of witchcraft, and in the veracity of historic accounts of its influence. What changed, in Davies's view, was that they no longer thought that the misfortunes that occurred in their own lives were caused by sorcery, nor suspected that some of their neighbours were witches. James Sharpe, in his study Instruments of Darkness, goes further and cites the same passage in Thompson's text as evidence that belief in witchcraft was unknown in England by the end of the nineteenth century.  Likewise, Geoffrey Scarre considers it to be evidence that benign and maleficent witchcraft were ‘slowly killed by disbelief’ and ‘near to extinction by the nineteenth century’.  In his article for the Athlone History of Witchcraft series, Roy Porter also cites the utterance as a symptom of the disenchantment of European popular culture; although, like Davies, he suggests that witchcraft belief gradually lost its currency over the course of the nineteenth and early twentieth centuries. 
There is, at least on the face of it, some evidence to support that view. Since at least the early modern period, drawing witches’ blood was believed to be a potent method to break their spells.  Acts of ritualized violence against supposed witches occurred even in Victorian England – allegedly at the vanguard of intellectual, as well as material, progress.  Thomas Hardy acknowledged as much in his novel The Return of the Native. The unconventional heroine Eustacia Vye is made a ‘martyr to superstition’ by a neighbour armed with a stocking-needle, who hoped thereby to put an end to a bewitching.  Real life witch-scratchings also found their way into the courts, where they caught the attention of journalists. In a pathfinding survey of the English local newspaper archive Owen Davies discovered over 200 ‘reverse witch-trials’, most of which date from the mid Victorian heyday of the provincial press.  Attacks on alleged witches did, however, begin to abate during the eighteen-seventies; although a trickle of cases continued until as late as the nineteen-fifties. 
It appears that no reported witch-scratchings occurred in Oxfordshire or south Warwickshire after 1875, although at least four had taken place during the previous twelve years.  Two of those incidents attracted much interest from journalists keen to appeal to the reading public's taste for sensational crime stories. One was also judged to be noteworthy because the assailant was a ‘respectable’ maltster – someone who was high enough up the social hierarchy to be expected to disown a ‘superstition’ like witchcraft.  The other case had such tragic consequences that The Times was compelled to moralize on the ‘deplorable’ mentality of the region's rural poor. In September 1875 at the village of Long Compton, barely ten miles from Flora Thompson's childhood home of Juniper Hill, an agricultural labourer named James Haywood savagely attacked one of his elderly neighbours, Ann Tennant, who died as a result. During his trial Haywood was unequivocal about his belief that his victim was the leader of a group of local witches, who used their occult powers to make him ill. It was perhaps because of his frankness that Haywood was judged insane, and thereby saved from the death penalty. 
The killing of Ann Tennant in 1875 was the last reported attack upon an alleged witch in Oxfordshire and south Warwickshire. However, the cessation of witch-scratching probably owed more to the growth of the rural police force than to waning popular belief in witchcraft. As Theodore Hoppen has noted, ‘the area into which the Victorian state spread its influence most dramatically was that of law and order’.  One way it did so was by replacing the part-time parish constables, whose loyalties often lay with their local communities rather than the letter of the law, with paid and independent police officers. Peel's legislation of 1829, and a further act of 1835, brought the new regime to London and most major towns.  Change was much slower to come to the countryside, however.  By 1853 still thirty of the fifty-two English and Welsh counties had failed to establish a countywide police force, including both Oxfordshire and Warwickshire.  That altered in 1856, when George Grey persuaded a reluctant house of commons to force all counties to follow the metropolitan example.  The result was more visible and effective law enforcement. The number of police in England and Wales trebled between 1851 and 1891, from one policeman to every c.1, 334 people in 1851, to one to every 731 in 1891. Overall criminality, too, appears to have gone into ‘a fairly steady decline until the end of the century’ – a notable exception being assaults upon policemen, which rose to a peak in the eighteen-seventies. 
The establishment of more effective rural policing seems to have curtailed attacks upon alleged witches. Doubtless many came to the same conclusion as the wheelwright from Lincolnshire, who in 1885 explained to the folklorist R. M. Healey that he had refrained from drawing the blood of a woman who had ‘overlooked’ him because of the threat of a fine from the local magistrates.  Previously many witch-scratchers had actually anticipated that the courts would take their side, and prosecute their supernatural persecutors instead.  James Haywood of Long Compton evidently expected as much, when he attempted to prove, by biblical quotation and folkloric technique, to his incredulous gaolers that his victim was indeed a true witch.  Such an expectation was not as ridiculous as it first sounds. Of course, since the amendments to the Witchcraft Act in 1736 neither parsons nor magistrates had been sympathetic to attempts to identify or punish witches. Some justices of the peace were so astounded that adults could believe such things caused their ills that they occasionally sent people who thought so for assessment at the local lunatic asylum.  Nonetheless the official disavowal of witchcraft had often not extended as far as the old-style parish constable. During the early nineteenth century there were instances – such as at Great Paxton in Cambridgeshire in 1808, and Wickham Skeith in Suffolk in 1825 – of parish constables standing idly by while their neighbours ‘mobbed’ or ‘swum’ reputed witches.  At Abergavenny in 1827 the parish constable himself took a leading role in the proceedings. 
Oxfordshire and Warwickshire saw a similar tolerance of witchcraft vigilantes, and other plebeian favourites like prize-fighting, during the first half of the nineteenth century.  Judicial inertia was evident in 1837, for example, when a contributor to the Oxford City and County Chronicle complained that ‘An inoffensive widow’ from his village had for some months been ‘exposed to perpetual annoyance and prosecution’ by her neighbours, who believed that she was responsible for bewitching a mentally deranged local girl. Not only had the widow been made to face up to ‘Cunning men and pretenders of every variety’ who had been commissioned by the girl's family, she had also ‘been prevailed upon … to be blooded by the poor girl in the arm with a pin, which has occasioned a malignant sore not yet healed’.  The editor of the Chronicle put his ‘trust [in] the Clergy and influential persons in the neighbourhood … to put a stop to such disgraceful proceedings’. It is little wonder that he called upon the officers of the church and the county bench, rather than the local parish constable who had obviously done nothing to stop the escalating persecution. With the introduction of a professional police force, no longer bound by neighbourhood associations or loyalties, such appeals would no longer be necessary. Following some well-publicized prosecutions in the eighteen-sixties and seventies, acts of ritualized violence against supposed witches ceased to occur, doubtless because the would-be perpetrators came to realize that the authorities were not on their side.
Belief in witchcraft continued to be widespread in Oxfordshire and Warwickshire, long after the district's last recorded witch-scratching occurred in 1875. On that point, virtually all of the local folklorists were in agreement. One reported in 1893: ‘In the course of my Warwickshire rambles in out-of-the-way localities, I have everywhere found amongst the lower classes a lingering belief in witchcraft’.  Neither was it only in isolated hamlets and bucolic villages that credence was given to the reality of sorcery. As the same scholar noted, ‘but a few months since’ in the market town of Stratford-upon-Avon, ‘a wretched woman who sold eels from the river … fell down dead by her own door-step, and was said by her neighbours to have been carried off this suddenly by the devil because she was a witch’.  Another investigator warned that in some villages, even at the turn of the twentieth century, witchcraft belief ‘still survives, in spite of the march of education and civilising influences, with a strength that is almost dangerous’.  Eighteen years later he reiterated that the belief ‘was still strong’, and speculated that it had been intensified ‘owing to the dreadful influences of the war upon the superstitious mind’.  As late as 1929 the folklorist of ‘Shakespeare Land’, J. Harvey Bloom, contended that ‘Witchcraft still exists or is believed to do so’.  Bloom's sentiments were echoed, in 1930, by another student of south Warwickshire, who cautioned his readers: ‘Lest any should imagine that witchcraft belongs to the middle ages only, and is now a thing of the past, it is worth mentioning that this is not quite so, and that belief in witches and their power still exists’. 
The folklorists’ observations about the survival of such apparent credulity, throughout the Victorian period and well into the twentieth century, are corroborated by the autobiographical sources. The author Evelyn Mordaunt, who lived in north Oxfordshire during the eighteen-eighties and nineties, encountered those attitudes among her family's servants. In her memoirs she reminisced about one – ‘old Job Ilkley’ – who ‘believed implicitly in ghosts … [and] also believed in witches’. His convictions were revealed when she proposed a sketching trip to a village on the south Warwickshire border (Long Compton) and ‘he begged me, with tears in his eyes, not to go there, for it was “fair swarmin’ wid witches”’.  Indeed he was right about Long Compton, which was known as ‘the very heart of witches’ country’.  Far from decrying their home's reputation, the natives of Long Compton were as persuaded as anyone that this was the case. A former resident recalled that during the first decades of the twentieth century most villagers were ‘Powerful’ interested in witches … If anybody were bad or a chimney pot fell off or some dire calamity befell, the witches were always to blame’.  Those remarks suggest that we are not dealing only with an abstract belief in the possibility of witchcraft, but that people continued to blame witchcraft for specific misfortunes – particularly accidents and illnesses – that blighted their own lives.
Long Compton had the reputation as the preeminent home of witchcraft, but belief in sorcery was found almost everywhere. The region's Gypsies, it was noted in 1912, had a considerable apprehension about some among their community who were credited with ‘occult powers’.  It seems that folklorists were also correct to suggest that accusations of witchcraft were made in bustling market towns as well as in the countryside. George Hillen's memoir of life in Stratford-upon-Avon, The Dillen, records both his own and his neighbours’ belief in ghosts, the innate healing powers of people with ‘the gift’, and witchcraft.  Hillen was born in the town in 1878, and raised by his mother's aunt – a lodging-house keeper, who had to bear occasional allegations that she was an ‘ole witch’: ‘folks [used to] say she rode up the fields a-night, on a besom [broom]’.  Similarly, in the village of Ewelme in Oxfordshire, ‘the power of the witch to work evil was still firmly believed in’ during the first decades of the twentieth century. 
Witchcraft belief continued to inspire forms of persecution even after the law had put an end to witch-scratchings and witch-mobbings. One legal but nonetheless effective variety of harassment was ostracism. That was used in one Oxfordshire village in 1884, against a young man who was thought to have ‘procured the “ill-wishing” of his mother’. With the exception of the rector's wife, ‘All the women in the village took up the matter warmly’, endorsing the diagnosis that witchcraft was to blame for the fits that had recently afflicted their neighbour, and driving the allegedly responsible party out of the community.  Very similar events, where witchcraft was blamed for causing fits or seizures, and many among the local community supported the diagnosis, occurred in the north Oxfordshire village of Souldern in 1876, and the market town of Deddington in 1869.  Those three episodes show that witchcraft was still being blamed for causing misfortunes after the cessation of the practice of witch-scratching. They also bear witness to how widespread the belief was; neighbours came out in force to endorse the idea that the afflicted were indeed bewitched. The strength of community feeling on those occasions was such that it commanded the anxious attention of the district's poor law guardians and journalists.
Drawing the blood of a witch was not the only way to break the power of spells. Those who believed themselves bewitched usually relied upon a multitude of techniques, supposed to provide relief without the risk of a direct confrontation with the witch or the magistrates. The most common of those was to nail horseshoes on to the doors of both domestic and agricultural premises ‘to break the charm of the witch as she approaches’.  They were still being used for that purpose at the close of the nineteenth century; so much so that, as one folklorist noted in 1898, ‘Any one visiting the hamlets of Oxfordshire can hardly fail to notice the numerous horse-shoes affixed to the picturesque thatched-roof cottages’. 
The horseshoe was only one of many objects that were thought to be potent against witchcraft. Crosses (made out of wood), knives or even opened scissors were positioned near doors, windows and chimneys, seemingly in the belief that witchcraft's evil influence tended to seep into a house through those locations that were not entirely sealed off from the outside world.  For a similar purpose shoes – then expensive commodities – were placed in roofs, mummified cats were bricked up behind walls, and witch-balls were hung from rafters.  Some people even sought to shield their own bodies, as well as their houses, from witchcraft. One of many ways to do that was by carrying renowned counter-witchcraft substances, such as the foot of a goose or certain herbs, in a small bag worn next to the skin.  That ‘secret way of believing’, as one autobiographer described it, was still in use in the Oxfordshire village of Ducklington during the nineteen-twenties. 
Traditionally the most agonized bewitched had sought the assistance of professional occultists known as ‘cunning folk’, about whom Thompson said absolutely nothing in Lark Rise to Candleford. Cunning folk offered a variety of services, including archaic medical therapies, finding lost or stolen property, love magic, and insights into the future. What most distinguished them, however, were their counter witchcraft services, which purported to identify the responsible witch and repel their baleful influence. 
Although cunning folk had always been controversial, attitudes towards them hardened during the nineteenth century, as the medical establishment, the judiciary, and the middle and upper classes added their voices to a growing chorus of criticism. Previously most educated men had ridiculed cunning folk for being ineffectual magicians who wasted their time on false learning and useless efforts to conjure spirits. The antiquarian John Dunkin, in his jovial biography of the Oxfordshire cunning man James Jagger, articulated the tail end of that attitude. The book describes how an ostentatious shoemaker ‘got himself laughed at’ by studying astrology and magic, and trying to use his learning ‘to raise the Devil’.  As the nineteenth century progressed that sort of aloof mockery gave way to calls for state-enforced suppression. Legislation of 1815, 1824 and 1858 made it easier for the courts to prosecute quacks like cunning folk, and reflected the growing feeling among lawmakers, professionals and civic bodies that cunning folk were charlatans representing a disreputable trade.  The Lancet identified that consensus when it wrote of cunning folk in 1864: ‘It is satisfactory to find a determination abroad to crush this system of imposition, as well as punish the imposters’. 
Attempts to suppress cunning folk were as active in Oxfordshire and Warwickshire as anywhere else, though they bore little fruit before the eighteen-sixties. The first sign of a hardening of attitudes among the middle and upper classes came in 1819 in the form of a response to an advertisement recently published in Jackson's Oxford Journal – the most widely circulated and respected of the region's newspapers. The offending advert had declared in the politest terms that: ‘John Manning, of Hailey, near Witney, late partner for many years with his father, William Manning, begs to inform the community at large that he has opened a shop at the above place, for the inspection of Urine, and vending of Medicines’.  The advertisement first appeared in late June 1819, and was reprinted throughout July, until one reproachful reader wrote to the editor to inform him of the true nature of John Manning's enterprise. The correspondent was W. J. Palmer, the rector of the small north Oxfordshire parish of Mixbury, a stone's throw from Flora Thompson's birthplace of Juniper Hill. The Revd. Palmer railed against ‘the villainy’ of those, like Manning, who made a living from dubious medical techniques such as urine scrying – ‘[a] shameful imposition practised upon the credulity of our poorer neighbours, not at the expense of their pockets only, but of their health and very existence also’.  Palmer's familiarity with the ill effects of such quackery came not from a personal knowledge of John Manning, or of his father and tutor William Manning. Rather he knew his brother Edward, who had recently set up in the same line of business in the Northamptonshire village of Croughton, on the Oxfordshire border, and only a short walk from Juniper Hill. The churchman had another reason for his disapprobation of the Mannings – a reason that he did not mention in his letter to Jackson's Oxford Journal, perhaps for fear of giving the enterprising family even more publicity. The ‘water-doctor’ of Croughton was also known locally as ‘Wizard Manning’ because he – like his father and his brother – was a cunning man. 
Edward Manning was twenty-seven when he set up his practice in 1819. Despite attracting the ire of local clergymen and magistrates, he had an extremely long career, which lasted for over half a century. Manning attempted to court respectability by advertising himself as a ‘surgeon’ or ‘doctor’ in the various local gazetteers that were produced during the period.  His efforts were in vain, however, as officials and journalists were increasingly scathing about his activities. At an inquest held at Leckhampsted (Buckinghamshire) in July 1830 into the death of an agricultural labourer, it was heard that the deceased, who had been ill for some weeks coughing up blood, had dismissed the services of a qualified surgeon in favour of taking a fifteen-mile ride on horseback to consult Mr. Manning. Needless to say, the ‘water-doctor’ had assured his patient that his medicines ‘would certainly stop the spitting of blood’; but the price for such an efficacious cure would be ‘all the money he had’ – that being 6s.  The ‘stuff’ – as the medicines were derisorily described at the inquest – failed to induce a cure, and the afflicted promptly died, giving the gentleman coroner of Buckinghamshire occasion to deplore ‘the folly of applying to such ignorant pretenders to physic as water doctors … every one of whom is liable, under an Act of Parliament’.
There were indeed cunning folk from the region who were prosecuted for taking their clients’ money under false pretences, although Wizard Manning was not one of them; the worst he had to endure was condemnation by the judiciary and the press.  In 1855 an inquest into the death of a woman from Kirtlington who had been to him for treatment ended with the familiar attack upon ‘quacks’.  The high point of Manning's infamy occurred in 1875, during the trial of James Haywood for the murder of the supposed witch Ann Tennant. It was revealed in court that Haywood had lost faith in the secular medical establishment's ability to heal him, and had gone instead to consult the ‘wise man’, who had confirmed that Haywood's illness was caused by witchcraft and told him the names of those responsible. For this encouragement of destructive superstition Manning earned the censure not only of Judge Baron Bramwell, but also of The Times itself, which ‘share[d] the regret of James Haywood's counsel that the “wise man from near Banbury” was not on trial for practising on his neighbours’ folly’. 
That denigration was to little effect, not least because the wizard, after a career lasting fifty-seven years, died a month later in January 1876.  His business had been an unusually lucrative one, even for cunning folk who sometimes made vast amounts of money from their specialist occupation.  Manning sunk much of his earnings into land, and by 1871 owned 170 acres in his local area and employed four servants, a housekeeper and a number of outdoor labourers.  The disparaging pronouncements of magistrates, coroners and journalists appear to have done little to tarnish his reputation among his patrons, who insisted that his prescriptions did ‘a great deal of good’, and kept his reputation as an enemy of witches alive for decades after his death. 
The attempts to suppress cunning folk were only partially successful. Around Oxfordshire and Warwickshire, as elsewhere, there does seem to have been a reduction in their number during the later nineteenth and early twentieth centuries.  They were not entirely stamped out, however; not least because a substantial proportion of the populace continued to have faith in their abilities. In 1912 two authors for the Journal of the Gypsy Lore Society reported that ‘old Josh Loveridge’ from Towcester in Northamptonshire was still doing a profitable trade as a cunning man.  The reason for his success was that he was patronized by ‘all the Gypsies’ who travelled through Berkshire, Oxfordshire and south Warwickshire. A Gypsy who lost his horse would not take the trouble to look for it himself, ‘but goes straight to Joshua and explains his difficulty to him’. Other Gypsies, perhaps fearing that they were bewitched, had taken to ‘paying him a weekly sum when … [their] luck is bad’. Such was the faith in the wizard's power that: ‘even incredulous Gypsies have a fear of offending Joshua, and declare they would give him their last shilling, if he asked for it sooner than incur his displeasure’.
Most of those individuals renowned for ‘black’ witchcraft did not, as far as one can tell, make any attempt to manipulate occult forces to do harm. There were, however, a few people who publicly claimed to possess the power to cast evil spells. In most cases those boasts appear to have been little more than a cynical begging strategy. As one Warwickshire folklorist noted about an elderly woman with a ‘reputation for witchcraft’: ‘There is no question that she greatly presumed upon her eerie reputation in order to get round the village mumping, as she called it, which … means that she used to call round on the people and collect from them oddments, a bit of tea from one and a few scraps of bread from another, and so on’. 
A few people had a reputation for witchcraft that was altogether more deserved. For a fee they offered to attack their clients’ enemies with maleficent magic. One scholar of rustic speech heard a rumour that there was a man who lived near the north Oxfordshire town of Banbury in the eighteen-eighties ‘who made his living by making little images to be stuck with pins for witchcraft purposes’.  Stories of people who ‘used to make an image of their enemy and stick pins into it, after the manner of old witches’ were also circulated elsewhere in the region.  Neither were those narratives pure invention or fantasy. Writing in 1930, one Warwickshire folklorist claimed that he could ‘introduce anyone who disbelieves that it is possible for such a mediaeval rite to be practised in this twentieth century to an old lady who will, for a fee, ill wish one's enemy with suitable incantations’.  The technique used by the would-be witch to ‘ill-wish’ her patrons’ antagonists was not image magic. Instead she conducted a ceremony in which she ‘muttered curses’, while burning ‘nuts on which the name of an enemy has been scratched’ over her fire. That she continued to make a living by such means, as late as 1930, suggests that Flora Thompson did not describe the reality of late Victorian and Edwardian rural life when she wrote in Candleford Green that ‘A few innocent charms and superstitions were all that remained of magic’. 
We can see from evidence cited above that, in nineteenth- and early twentieth-century Oxfordshire and Warwickshire, many people continued to believe in the abstract reality of witchcraft. Not only that, they sometimes concluded that accidents, illnesses and other misfortunes that blighted their own lives were caused by particular witches, who resided in the neighbourhood. And yet, despite the widespread private concern about the apparent influence of maleficent witchcraft, in public discourse most individuals, like the protagonist's mother in Lark Rise to Candleford, repeated the mantra that witches ‘seem to have all died out’.  Evidently, that duplicity demands an explanation.
In explaining this quandary some works of later twentieth-century anthropology are most helpful. They reveal that a mismatch between witchcraft beliefs and public testimony is nothing new. The French ethnographer Jeanne Favret-Saada encountered something similar during her fieldwork in the Normandy Bocage during the nineteenth-sixties and seventies. According to the theory of witchcraft that was current in that region, ‘spoken words are power, and not information’.  The seemingly innocuous act of talking about witchcraft was considered to be extremely dangerous, because it was an activity that was sure to inspire a local witch to focus their malevolence on the speaker.  That sentiment was expressed in common parlance by the phrase ‘the less one talks, the less one is caught’.  Even listening to stories about witchcraft was thought to have the potential to change the course of an individual's fate, by somehow agitating a witch to aim his spells at the curious inquirer.  As Favret-Saada summarized, her informants thought that ‘it was very dangerous to talk of spells, and even more dangerous to try to understand them’.  For that reason, those who believed that witchcraft posed a genuine threat were, in most cases, reluctant to discuss it, unless it was associated with parties who were safely dead.
The inhabitants of the Normandy Bocage were by no means unique in their aversion to talking about contemporary witchcraft. On the contrary, the notion appears to be quite widespread, and has been documented by scholars of geographically disparate cultures. Ann Ross, in her 1976 study of the folklore of the Scottish highlands, reiterated many times that witchcraft ‘is a subject that people are often unwilling to speak about, but there is no doubt that belief in powers to work good and evil is still fairly widespread’.  Further afield, on the border between Mexico and the United States, the Pueblo also believe that talking about witches will tempt their malice. One anthropologist noted in 1976 how that belief ensured his fieldwork was all the more difficult: ‘Information on Pueblo witchcraft is more uncertain and in smaller supply than is the case for most other aspects of the culture. This is easily understood when it is realised that the Indians are loath to discuss the subject in detail for fear that supernatural powers may somehow retaliate’.  Scholars of Japanese, north Indian, east African and Nepalese cultures have also recorded that their informants harbour the same belief about the supernatural significance of speech, and the dangers that follow from openly discussing contemporary witchcraft.  Nancy Levine, for instance, noted that the Nyinba people of Tibetan Nepal were ‘reluctant to discuss witchcraft (ngan), because, they said, the witches might retaliate against them for doing so’.  Likewise a recent study of Indian sorcery has described how ‘the fear that the witch might afflict a person for levelling an accusation against her … undoubtedly prevent[s] many people from voicing their superstitions’. 
There is good reason to think that similar beliefs were prevalent in Victorian and Edwardian England. A strong aversion to talking about witchcraft was certainly current in Oxfordshire and Warwickshire. That sentiment was noticed by a number of social commentators and public officials, including the poor law relieving officer George Dew. In 1876 he recorded in his diary that a young woman from the village of Souldern had become ‘unwell & her mind perhaps perverted’ – the occurrence of which her neighbours and her father attributed to witchcraft.  Dew attempted to investigate the situation but found that he ‘could not persuade … [the girl's father] to tell me who it was [that bewitched his daughter], nor yet to disbelieve in witchcraft’. Elsewhere people were just as guarded. The Gypsies who travelled around the area were noted to be ‘generally reticent as to … [the] doings’ of the much respected ‘wizard’ of Towcester, Joshua Loveridge.  In Ewelme at the turn of the century, tales about old village witches were ‘whispered’, rather than recounted openly.  J. Harvey Bloom found that he could elicit such stories from his rustic confidants only ‘after much persuasion, and in much fear of the consequences’.  Reticence persisted among the rural population as late as the nineteen-sixties, as another student of south Warwickshire noted: ‘Many of the older people in this area today can recall tales of local witches being told by their parents. But they prefer not to repeat them’.  The observer speculated that the reason why people were so reluctant to repeat witchcraft anecdotes was that that they feared for the reputation of their ancestors. About that, it appears, he was wrong; as elsewhere, it was a concern for their personal safety that inspired a guarded silence. Mollie Harris, who was born in the Oxfordshire village of Ducklington during the nineteen-twenties, and spent some of her adult life documenting the habits of the locals, discovered that for herself when she tried to interrogate two of her elderly friends about the burial place of a local witch. ‘“Don't you ask neither of us wur tis”’, her informants coyly replied, ‘“cos’ we shan't tell you, cos that would bring us bad luck”’. 
Perhaps the fullest record of the reluctance to speak about witches can be found in the manuscript notes of Oxfordshire's most prolific folklorist, Percy Manning. In May 1904 Manning conducted an interview with Thomas Walker, a labourer from the village of Great Milton, situated about six miles south-east of Oxford. It is evident from the notes that Manning made during the interview that his chief objective was to gain information on witchcraft beliefs. However, this was no easy task because his informant sought to avoid speaking about witchcraft by being vague and changing the topic of conversation. 
The interview began with talk of a house at Haseley, which had been the scene of a notable death, and had since been locked up under the supervision of a caretaker. To Manning's frustration, his informant was not forthcoming with details. All Walker would say was that ‘Very queer noises came from it, and “things” were seen too’. As Manning wrote, ‘but all this was very vague; Walker wouldn't or couldn't be specific’. Pushed further Walker gave a few details about the house's caretaker: ‘This caretaker was Susan (“Sukey”) Buckland, the gipsey, who died about 15 years ago at Waterstock. People were afraid of her; Walker knew her; “twas said she was a witch”’. Manning was evidently intrigued by the story, and went on to talk about witches. Yet to his surprise – and from what he could tell, for no obvious reason – his informant changed the topic of conversation to a legend about an underground tunnel that ran beneath the village. As Manning wrote: ‘this was Walker's own transition’. Unwilling to give up the chase, he continued to press Walker for details about witchcraft, and was rewarded with two stories about a witch who, in the past, had lived in the village of Cuddesdon: ‘called mother Galloway, who had a spite against a farmer, and the farmer couldn't get his butter to come, although he churned all day. “Throw some pins in the milk” someone suggested. The butter then came all right, and mother Galloway was found scratched and torn all over’. Walker concluded his talk of witches with a brief account of an incident that had involved his son, when he lived in the village of Hatfield, just over the Berkshire border, in the Vale of the White Horse. The anecdote was, nevertheless, more vague and evasive than Walker's stories about Susan Buckland of Haseley, and mother Galloway of Cuddesdon. Indeed, Walker did not directly name ‘witchcraft’ as the agency at work; rather, he left that to be inferred by the folklorist. Moreover, he was careful not to identify the specific location where the episode had taken place, nor the names of the others who were involved: ‘[Walker's son] took out a team of horses in his cart, and at one particular place in the road, the horses “came out of the tackle [harness]” and left the wagon standing. He was very much frightened. Many others had the same thing happen at this place’. It is evident from Manning's notes that his informant was anxious not to speak about witchcraft. The interview was characterized by vagueness and seemingly needless digression; and Manning was able to discover little except two local stories about a witch who had troubled another village, some time ago in the past.
Folklorists often attributed their informants’ reluctance to discuss ‘superstitions’ to a fear of being thought credulous. Charlotte Burne, in the first edition of the Folklore Society's Handbook of Folklore (1890), articulated a commonly held view among her colleagues when she explained that: ‘It is the first instinct of the folk to deny all knowledge of superstitious practice, out-of-the-way customs, or curious legends. They are afraid of being laughed at’.  Others misconstrued it as the result of mistrust, if not hatred, engendered by the class system.  Active as both of those impulses were in folklore collection, it was nonetheless the case that the aversion to talking about witchcraft was inspired by a different sentiment. The findings of another south Warwickshire folklorist, F. W. Bennett, also support that conclusion. Bennett was no brash or aloof outsider in the villages that he studied during the nineteen-twenties. He had spent years befriending ‘agricultural labourers of the old school’, usually in the congenial atmosphere of the kitchen of a ‘country inn, well off the main road’.  In spite of that, he too noticed the ‘extreme reticence of country folk’ on the subject of ‘superstitions’.  One telling feature of that reticence was, however, that it was selective rather than general. The villagers, he noted, ‘are extremely free in the matter of superstitions as to good or bad omens’. They were, for example, quite open about their belief that it was ‘bad luck’ to see a new moon through glass, to turn bedding on a Sunday, or to transplant parsley rather than grow it anew from seed. Likewise they were only too happy to explain that some sort of supernaturally derived disaster was likely to follow from cutting one's nails on a Friday or Sunday, hearing a dog howl at night, or placing three lighted candles at a table.  Bennett's crucial observation was that achieving candour about those sorts of superstitions did not mean that the folklorist had gained ‘admission into the inner confidence’ of his or her informants. As he noted, it was only necessary for the interrogator to ‘broach the subject of the supernatural’ – meaning ghosts and witchcraft – ‘in order to find out one's mistake’. For when the particular superstitions of ghosts and witchcraft were raised informants either let ‘loose a spate of the most impossible yarns, or [else] there is a frigid silence’. Bennett did not attempt to explain why his informants were selectively taciturn about their beliefs, though he was certain that ‘despite all their reticence these people are profoundly superstitious’. We can be sure, however, that it was not shyness or animosity that motivated that differentiated silence. Informants were, after all, happy to invite ridicule by openly conceding their belief in ‘superstitions’ about omens and luck. When people skirted around the subject of contemporary witchcraft they did so not because they were concerned for their reputations in the eyes of middle-class antiquarians, but because they feared something much more consequential to their own well-being.
Given that they wanted to avoid the dangerous activity of discussing witches who lived in the present, why did folklorists’ informants go to the trouble of recounting tales about witches from generations past? It certainly would have been easier for believers in sorcery to say less about the subject by maintaining that witchcraft was imaginary, but it would have been barely less hazardous. Just as it was believed that speaking candidly about witches who were still living would anger those evil ones to vengeance, so too was it believed that denying outright the abstract reality of witchcraft might also invite a supernatural assault. As the minister of Bicester congregational chapel noted, with regret, in 1848: ‘Witches and ghosts are, in the firm opinion of the sweating classes, a terrible reality – a reality with which it is most dangerous to trifle. Those who make free to call in question the witch's power and the ghost's visible appearance, run the risk of being visited by these wandering phantoms’.  The idea also came up in the local tales of the supernatural, where those who denied the reality of witches were destined from that point on to meet with them.  When it came to discussing the supernatural, the hubris of scepticism was believed to be as dangerous as plain speaking in tempting fate to deal an undesirable hand. As a consequence, those who wanted to evade the witches’ malevolence trod a fine line between frankness and dishonesty.
One reason that people told stories about witchcraft from the past was to avoid talking about witches who were believed to be operating in the present. However, it would be wrong to think that the tales of witches from the previous generation were invented for the benefit of inquisitive folklorists. On the contrary, legends of deceased witches were being recounted in both the countryside and in the towns long before the arrival of the folklorists in the later nineteenth century.  That many different informants independently told folklorists the same legends, albeit with some small variations in detail, proves that local tales about reputed witches were substantially stable, and widely circulated.  The witches who were named in the narratives were, moreover, people who really had resided in the district. A perusal through the minutes of the parish vestry meetings soon demonstrates that ‘the Standlake Witch’ Sarah Wade, who one folklorist was told about in the nineteen-twenties, had indeed lived in that Oxfordshire village during the eighteen-twenties.  Likewise, the reputed witch of Headington, Miriam Russell, who was the subject of a witchcraft story that was collected by a folklorist in 1897, was buried at the parish church in 1845.  Moreover, folklorists’ informants were all too often ready to vouch for the truth of the tales. They did that either by saying that they ‘knew the people involved’,  or by invoking the great authority of the King James Bible, with its story about Samuel, Saul and the ‘Witch of Endor’, to put the reality of witchcraft beyond doubt.  The local witch stories were, thus, the collective memory of a close-knit rural society; and their circulation ensured that a person's, and by implication a family's, reputation for witchcraft stayed with them for generations. 
Thompson did accurately record a common remark when she made one of her characters say that witches ‘all seem to have died out. There haven't been any in my time’.  However, it would be wrong to follow Thompson and several other scholars in interpreting remarks like that as evidence that witchcraft belief was declining. Instead, those utterances were part of a strategy for avoiding speaking candidly about witches who were still living. The reluctance to talk about them stemmed from a fear that, far from having disappeared, witches posed a genuine threat – especially to anyone who highlighted their existence. The contention that they ‘all seem to have died out’ ought, therefore, to be seen as evidence of the continuing vitality of belief in maleficent witchcraft, rather than of its decline.
That is not to suggest that witchcraft beliefs and practices in Oxfordshire and south Warwickshire remained unchanged throughout the nineteenth and early twentieth centuries. The cessation of witch-scratchings and the decline of cunning folk were significant changes in plebeian mores. In the case of the former, if not the latter, the primary cause was the increasingly effective implementation of the law by the burgeoning rural police force. That much is evident from the fact that witchcraft stories continued to be created and circulated for almost a century after the last assault upon a supposed witch, and that counter-witchcraft practices and maleficent magic were still in use during the early decades of the twentieth century. Belief in witchcraft continued to be practical as well as abstract, with believers still suspecting their neighbours and using various methods to protect themselves from supernatural aggression.
Particular words, when uttered in certain contexts, were believed to have the power to unleash – or, at least, unsettle – supernatural forces. That was true not only with respect to witchcraft, but also across the entire spectrum of plebeian superstition. Folk stories recounted how the Almighty had blighted infidels who publicly professed their unbelief.  Few people cared to refer directly to the Devil, who was always known by his nicknames of ‘Old Nick’ and ‘Old Scrat’.  Likewise farmers and shepherds were often unwilling to say how many lambs or calves they expected from their herds, for fear of tempting fate.  Some Victorians and Edwardians were, therefore, uneasy about discussing the supernatural in a straightforward way, not because they were embarrassed about their beliefs, but because they had a vague sense that it was treacherous to speak plainly about them. Acknowledging as much is essential if we are truly to understand the subtleties of plebeian religion, and correctly interpret the meaning of popular religious discourse.
In Thompson's imaginary Oxfordshire, education had expunged the ‘old dark superstitions’, like witchcraft, from popular culture.  Her portrayal of an incredulous rural labouring class was probably motivated by a wish to idealize the rural past. As Ronald Hutton has observed, within later nineteenth- and twentieth-century literary culture there was a ‘tremendous idealisation of rural England’.  Lark Rise to Candleford was a contribution to that movement, but it was a conservative contribution, which in some ways harked back to nineteenth-century values of respectability and intellectual self-improvement. Like Victorian newspapers and a number of authors from that era, Thompson regarded belief in witchcraft to be a crude and destructive superstition that ought to be eradicated by education.  Perhaps that is why she expunged it from her imagined village.