Exploring the beliefs of English people during the fifth to ninth centuries


what's new?

introductory articles

in-depth articles


join the email update list

Anglo-Saxon Twilight only

Anglo-Saxon Twilight
is currently sponsored by
Heart of Albion Press
publishers of Bill Griffiths'
User Friendly Dictionary of Old English

Who were the landwights?

We are perhaps the last generation to have direct access to the oral tradition of local communities who have a deep attachment to a land made of what is visible and invisible, beliefs and rituals attached to natural and humanly altered rocks which meaningfulness was reinvented in the context of a traditional modus vivendi. With the advent of literacy and the profound changes implemented in the rural world, the dilution of a communal identity and practices like story telling have contributed to the rapid loss of an ancient imaginary world, the moors and the legends, the spirits in the rocks and the rivers… As the generation of the 'eldest' leaves, stories on ancient sites are progressively being replaced by scientific history in the discourse that reaches local communities.
(Alves 2012: 270 fn 1)

The Old English word landwight seems to correspond to our phrase 'spirits of place' and the Latin genii loci. In other words, it is a collective term for a variety of what are often thought of as supernatural or Otherworldly entities.

Actually, for reasons I have discussed elsewhere (see The potency of leeks and the spirit of alcohol), pre-conversion Anglo-Saxons seemed not to have thought in terms of either the 'supernatural' or of Otherworlds. These are concepts which slowly take root from about the tenth century onwards as Christian worldviews permeate popular thinking. So far as we can tell from Old English literature the landwights were not Otherwordly but instead were immanent – that is, they were part of this world, albeit perhaps not always so visible or tangible as other aspects. The more appropriate term is 'preternatural'.

The Anglo-Saxons recognised many different types of landwight and gave them names such as aelf, thyrs, eoten, puca, scucca, dwearg and maere (with any number of variant spellings). But this is in itself confusing as aelf – which gives us the modern word 'elf' – did not simply denote diminutive hominids but was also a collective term, in the same way as landwight. But aelf and landwight did not denote the same 'collection' of entities.

Getting confused? It gets worse! Think of the modern word 'fairy' and you will probably think of twee, gossamer-winged entities most commonly found in children's books. However, if you're a well-informed British folklorist you will think of entities as large as an adult human, definitely without wings, sometimes helpful but generally not to be trusted, and who may or may not be witches rather than fairies. And then there's the problem of books of 'fairy stories' which, in large part, do not have any fairies in them – though they are all tales of enchantment and magical happenings.

hollow hills


Hollow hills – as depicted in tradtional lore (top)
and the 1990s BBC children's television series Teletubbies (bottom).

Aelf has many of the senses of 'fairy' in the term 'fairy stories'. Aelfhame – 'elf home' if you like – is a realm where enchantment and magic take place. It is similar to 'fairy land', especially a fairy land thought of as being a real place under a hollow hill.

However in the mind of an Anglo-Saxon the word aelf included entities which are alien to modern concepts of fairy. Aelf included the os, regen and god. All these are collective names for deities – although we should be thinking more of local spirits of place rather than the universal deities of Classical and Christian worldviews. Although, yes the Old English word god was 'hijacked' – with a considerable shift of meaning – to be the name of the Christian God. By extension 'god' became the collective name for male deities – although in Old English the word os has exactly this sense, while god did not!

All things bright and other-than-human


The only known Anglo-Saxon illustration of aelfe.

At the risk of over-simplifying things, aelf denotes all the good and 'bright' aspects of such other-than-human entities. There is even an Old English word aelfscinu which means 'elf shine'. While scin, like 'glamour' was used to describe a beautiful woman – 'not simply beautiful, but perilously so' – its main usage was to denote the deceptive appearance of anything 'paranormal'. While the evidence is ambiguous, early sermons and such like seem to regard biblical angels as aelf (see elf shine article).

The related word scinna specifically refers to a ghost or apparition, although more commonly the word used for ghosts is grîma – which, to cause total confusion to later scholars, is also a common personal name! So, whatever aelf may have been, they were not confused with ghosts.

So if that's the meaning of aelf more-or-less sorted then what about the meaning of landwights? Again, at the risk of over-simplification, if aelf denotes the 'goodies', then landwights denotes the 'baddies'. Or, if not bad all the time, much less trustworthy than the only-sometimes deceptive aelfe (the plural of aelf is aelfe, not aelfs or aelfen).

Never name the fairies

One reason that we may have difficulty understanding exactly who is being referred to by the word aelfe is that there may have been a cultural taboo on using the 'proper name'. Among many rural people in Europe to this day it would be considered 'unlucky' – and unnecessarily risky – to refer to the fairies by name. Euphemisms such as 'the wee folk', 'the Gentry', 'the Hidden Company' and in Welsh Y Mamau ('The Mothers') are used instead. (See The Mothers article.)

There is no direct evidence that Anglo-Saxons used such ambiguities intentionally. However Old English literature abounds with kennings – metaphorical by-names – so any such evidence would be hard to recognise. Indeed, if such a taboo existed and was consistently maintained then it is hard to envisage what clear evidence of such name-avoidance might be.

The types of wights

Many of these wights live on in modern English words. Even the word 'wight' is found in contemporary dictionaries, with the sense of 'living being or creature' – although most usage of the word is as an archaic synonym for a ghost or such like.

Giants of the land

The word thyrs in contrast has no heirs in modern English. Instead the Old French word geant took it's place – from which our word 'giant' derives. The Old English word eoton lived on until the sixteenth century as 'ettin'. Both are related to the Old Norse word jotunn. If there was ever a difference between a thyrs and an eoton then sadly the surviving Old English literature does not make this clear. We must think of them all as giants.

Carn Ingli

The 'sleeping goddess' near the summit of Carn Ingli, Pembrokeshire.

Beacon Hill

The giant's face on the summit of Beacon Hill, Charnwood Forest, Leicestershire.


The Long Man of Wilmington, Sussex.

Anglo-Saxon belt buckle from Finglesham

Anglo-Saxon belt buckle from Finglesham, Kent.
His pose suggests that the Wilmington giant may also have Anglo-Saxon origins.

The natural simulacra of the landscape – such as the recumbent 'goddess' near the top of Carn Ingli in Pembrokeshire and the giant's head on the summit of Beacon Hill in the midst of Charnwood Forest, Leicestershire – were presumably called by one or the other name.

If so, were the man-made figures carved into chalk – and the Long Man of Wilmington is sufficiently similar to Anglo-Saxon decorative metalwork to have been created back then – known by same name, or by the other name?

The ancestors of Puck

The Old English puca evolve into the various tales of Puck or Pook, who in turn becomes indistinguishable from Robin Goodfellow. Despite a substantial number of tales about Puck and Robin Goodfellow surviving from the early modern period, sadly most British people are unaware of this. They were both tricksters in the same fashion as Coyote, Brer Rabbit, Anansi and all the other New World tales worked to death by modern-day storytellers.

Puca were the most goblin-like of the landwights. Even in Shakespeare's A Midsummer Night's Dream, Puck is a mischievous fairy. We might also think of puca as 'impish'. Again, such sometimes-helpful, sometimes-troublesome human-like but diminutive beings abound in legends and tales, not least those written for children.

However the Old English words scucca or sceocca are often translated as 'goblin' even though the only survivor of this word in modern use 'Black Shuck' describes a phantom black dog.

Black dogs in folklore    cover
See Black dogs in folklore and Explore Phantom Black Dogs

Dwearg also denotes diminutive beings – they give us the modern word 'dwarves'. Quite how human – or not – the dweargen – were originally thought to be is a moot point. We now know that dwarfism is caused by several different genetic traits. However traditional folklore would suggest that such people would be deemed to 'fairy changelings' – or worse. Or maybe the Anglo-Saxons used the word to describe much less tangible entities, and the meaning simply transferred to denote human dwarves.

Beware the Hobbits

One thing we must be careful not to do is think of dwarves and elves in the same way as J.R.R. Tolkein's elaborately-constructed realms of 'dark elves', 'light elves' and industrious dwarves. Yes Tolkein was a fine Anglo-Saxon scholar of this time, but most of his inspiration for Lord of the Rings came from Scandinavian sagas written down by Christian authors in the thirteenth century. Snorri conflated the elves with the dwarves and added a distinction between light and dartk elves which is not attested previously. Understandably Snorri's confusion has been passed down to many more recent writers.

Tolkien's tales reveal his own wonderful imagination at work, in all its glory. But they reveal nothing about what Anglo-Saxons may have thought aelfe and dweargen were like – still less that they would have been akin to Tolkien's Hobbits and their ilk. Indeed the influence of Tolkein's books has served to thoroughly obscure 'real' Dark Age worldviews from anyone but the most rigorous academics.

Beware the mare

So far the entities making up the landwights and the aelfe are more-or-less human – some much more (the thyrs and the eoton) and some much less (the puca, dweargen and perhaps the aelfe too). But the maere seemingly never took human form. This is the modern word 'mare', denoting a female horse, and part of the word 'nightmare'.

'Nightmare' first appears in written English around the thirteenth century, although it was almost certainly in use orally before this. Back then it had the sense of an evil female spirit afflicting sleepers with a feeling of suffocation. If that sounds entirely paranormal, or just one type of unpleasant nightmare with no reality to it, then think again. Back in 1982 David Hufford published a splendid study, called The Terror That Comes in the Night: An Experience-Centered Study of Supernatural Assault Traditions (Hufford 1982). He convincingly argues that an experience known to folklorists as dreams in which sleeper is 'hag-ridden' are the result of a specific type of sleep paralysis.

However both the thirteenth century meaning and Hufford's research describe describe experiences where a female human entity visits the sleeper. They do not explain why the word maere – female but a horse not a human – was used.

Unusually we have a seventh century document where the word maere is glossed with the Latin incuba. This is from the verb incubo, incubare meaning 'to lie upon'. In later traditions it refers to a male demon – an incubus – who lies upon sleepers, especially women, in order to have sexual intercourse with them. Its female counterpart is the succubus.

In most of the accounts incubi and succubi were two manifestations of the same demonic beings, who could change gender at will. Repeated intercourse with an incubus or succubus might result in the deterioration of health, or even death, according to medieval legends. Although bowdlerised from Victorian versions of the Arthurian legends, Geoffrey of Monmouth attributes the conception of Merlin to his mother's sexual relations with an incubus. His mother was actually in good company – St Augustine was aware that incubi often made 'wicked assaults upon women' and St Thomas Aquinas, eight hundred years later, also discusses such attacks.

While the sense of being 'hag-ridden' is consistent with the broader sense of incuba although we have to assume then, as later, incuba was not necessarily gender-specific. But it does leave wide open why sources in Latin conceive of them as anthropomorphic – albeit shape-shifting – demons whereas the Old English is specifically zoomorphic. And we must bear in mind Merlin's mother – in other words, real children were attributed to such 'paranormal' encounters. One can only suspect that there was nothing paranormal about such conceptions, merely a deliberate attempt to conceal the identity of an all-too-human father.

The mare, the stallion and the horse

One day I will write a more extensive discussion about maere but suffice to say for now that this one type of landwight is the most revealing about pre-conversion Anglo-Saxon worldviews. We are looking all-but-directly at a time when Anglo-Saxons thought that their family lineage – their kin and, by extension the tutelary goddesses of that kinship (see The queen of the valley) – could be traced back to ancestors who took the form of horses.

We see this most clearly in the legend which names the first Anglo-Settlers in Britain as Hengest and Horsa. The names mean 'stallion' and 'horse'. Almost certainly the legends which have come down to us are highly distorted versions of pagan 'foundation myths' which straddle the migrations from northern Europe. Whether Hengest and Horsa were tutelary deities – I suspect they were, but of course have no way of proving this – then one horse deity was most certainly known by name, both in Europe and in Britain. And that was Epona, and she was a goddess – a mare.

iron age coin

While Epona was one of the more universal goddesses of pre-conversion Europe, she is least attested in Britain – there are only two certain inscriptions and only four more doubtful depictions. Were there local deities with similar attributes? Were the stylised horses on Iron Age coins some sort of 'divine protection' of the local equine divinity? Would an imported cult of Epona simply have not have been neeeded? Or were these local 'horse deities' regarded as tribal manifestations of a more universal deity? The parallel is the devotion shown to individual icons of 'Our Lady of X' by modern day Catholics while also fully recognising the same saint as a more 'universal' Blessed Virgin Mary.

Such an equine deity would have been a white horse (see The deities of the Anglo-Saxons). Indeed any white horse would have been regarded as a manifestation of her, or at least her local counterpart. My maternal grandmother, who grew up in south Leicestershire just before the First World War, would never speak of a 'white horse' – the tradition was to refer to them as 'greys' (even if they looked Persil white). This is of course the same reason one of the regiments of the monarch's household cavalry – who ride white horses – are known as 'The Greys'.

Irish medieval literature tells of the sacrifice of horses at the inauguration of kings – indeed Gerald of Wales recounts how one king was ritually 'married' to the horse, who represented the sovereignty of the land – in other words, the mare was a tutelary spirit. In Old English she would have been a maere, one of the landwights. And, as such, not just any old landwight, but one of the most preeminent in pre-conversion times. The evidence is of course skewed because most of what we know about maere comes to us from one or more centuries after the process of conversion as underway. And the Christian scribes were least likely to have any interest in, still less have a reason to record, anything about female horse deities.

The one thing those clerics did achieve was a taboo about eating horse meat which persists to this day. This is a peculiarly British avoidance, not shared in Germany – or even German-speaking Switzerland where I can attest that fohlenfleisch is served on a regular basis in works canteens.

Such culinary prohibitions are seemingly along way from the literal meaning of nightmares. But in a society when the primordial female goddess took the form of a horse then any unduly strange and disturbing nocturnal experience would be explained as an encounter with her. The only reason why we find this suggestion strange is that underlying 'assumptions' of modern thinking have shifted a long way from those of Anglo-Saxon antecedents. Only our customs and fragments of our language retain these 'fossils' of a long-distance past.

Here be dragons

The Anglo-Saxon Chronicle for 793 records for posterity how people of the time thought about dragons: 'If not flying through the air, dragons were to be found in barrows or caves.' Beowulf famously battles one in just such a place.

The Old English word is draco and appears in such place-names as Drakelow ('dragon mound' or 'dragon-protected mound'). There are examples in Bedfordhire, Derbyshire and Worcestershire. We must also assume that a dragon was watching over the treasure in places which take their name from 'hoard mound', such as Hurdlow in Derbyshire and Goldhard, near Tandridge in Surrey.

Seemingly distinct from dragons, at least until the later Anglo-Saxon era, are the wrym. This word denotes serpents and other reptiles. And equally distinct are the nicor, water monster. While nicor shared some of the same 'paranormal' attributes of the landwights they were of course not considered as 'land wights' simply because they were 'water wights' instead. And, for whatever taxonomic reasons now lost to us, draca and wyrm are seemingly not thought of as a type of landwight.

Where did the landwights live?

Considering how little we really know about the landwights, simply because the surviving literature is fairly sparse, we know considerably more about where they lived. Place-names incorporate a surprising number of aelf, thyrs, eoten, puca and dwearg.

Edward Smith has compiled a substantial list of examples of such place-names and made them available online.

See Edward Smith's Pagan and supranormal elements in English place-names

Since writing this article an up-to-date discussion of so-called 'supernatural' Anglo-Saxon place-names (including a complete list of such references in boundary charters) has been published as part of Sarah Semple's recent book Perceptions of the Prehistoric in Anglo-Saxon England (Semple 2013: 179–92; 261–81)

However Smith does not discuss the potential confusion between puca and pocca. The latter is a word for fallow deer and a number of the examples he gives for puca might instead be from pocca. Also places such as Elveden in Suffolk may not be the denu or valley of the elves but of the swans (Old English elfitu). Sadly the earliest written versions are often too recent to be reliable records of the original names.

The place-names scholar Carole Hough has discussed Pucklechurch in Gloucestershire and attempted to distinguish between pocca and pohha (both Old English words for fallow deer) and the diminutives pohhel and poccel (presumably refering to fawns) on the one hand and puca and pucel, both names for 'goblins', on the other hand. After nearly ten pages of erudite scholarship she conlcudes that this name probably originated as the church associated with a 'poccel', a young fallow deer – perhaps an aspect of the life of the founding saint (Hough 2012). There has been a counter-argument that fallow deer were not known in England until the twelfth century, but this would still allow for poccel to be incorporated in minor place-names. So possibly some 'Puck–' place-names originally evoked deer rather than goblins.

However this does not mean that landwights did not sometimes lend their names to places. As Susan Kilby has established in her as-yet unpublished thesis on place-names around Lakenheath, 'In the pre-Christian period, certain pits were associated by pagans with the entrance to the underworld, places where communication with supernatural beings might take place.'

Particularly suggestive are place-names such as thirspitt'thyrs (giants) pit'. Kilby also discusses the Old English word swalwe – the origin of the modern word 'swallow(hole)' – as originally having the sense of Ďabyssí. Swalwe is used in the the Anglo-Saxon poem Christ and Satan to refer to Hell.

Once her thesis is completed then I suspect more can be said about pits, hollows and holes in place-names. Some will be pragamatic descriptions, but just maybe others will provide an insight into pre-conversion beliefs about where some of the landwights lived.

Much more about aelfe

This article could not have been written without Alaric Hall's wonderful book Elves in Anglo-Saxon England: Matters of belief, health, gender and identity (Hall 2005). This is based on his PhD thesis which is available online (The Meanings of Elf, and Elves, in Medieval England).

There is much more about aelfe in Hall's writing than I have summarised here – not least the shift from masculine (although somewhat effeminate) aelfe of the eighth century to female aelfe of the eleventh century. Intriguingly, this is the time when Christian angels began to shift from being thought of as exclusively male.


copyright © Bob Trubshaw 2013–14


previous article     index of articles     next article
index of introductory articles
email comments and suggestions
join the email update list
what's new?

Anglo-Saxon Twilight only