TAKE a good look at this strange creature. Whatever can it be?

Now try to describe it in words: large head, big feet, strangely elongated nose, very large, fan-like ears. This is an elephant – or at least as good an approximation of one as could be managed by a 15th century craftsman from rural Suffolk who had never seen such a creature, but only heard it described.

To this honest artisan, skilful as he was, the elephant was probably no more or less real than the cockatrice, with its cock's head, dragon's wings and serpent's tail, that he carved with equally vivid care on another of the bench-ends in the parish church of St Nicholas in Denston. Or the man-eating manticore, with its bearded human head and lion's body, that one of his fellows created in bas-relief for the splendidly decorated wooden wall-plate that runs like coving all round the join of wall and ceiling of this wonderful church.

All these exotic beasts, and many more carved in profusion throughout the church, were familiar to the medieval parishioners from the bestiary, a 14th century bestseller which "explained" animals real and fabulous not in terms that we would recognise as "natural history", but in terms of their supposed relation to the scriptures.

We believe post-Darwin (or most of us do) that man is an animal: they believed that God had put animals on Earth to teach man how to lead godly lives. The bestiary, in its various versions and editions, was the handbook that explained the lessons.

In a number of other churches, such as Lakenheath and Fressingfield, the elephant appears looking more like the creature we know, but with a castle on its back. This is how it appears, influenced no doubt by ancient tales of Hannibal or distant ones from Persia and India, in many bestiaries: the copy available to the woodcarver in 15th century Denston obviously lacked this illustration.

It is intriguing to compare his vision with that on the high wall-plate in the grand church of Burwell, just over the county boundary in Cambridgeshire. This too, though complete with castle on its back, appears to be the result of carving from verbal description – or perhaps the dimensions of the wood led to the foreshortening, or squashing, of a bestiary image. Whatever the reason, Burwell's elephants are probably the nearest thing in East Anglia to crocodiles.

On its bench-ends alone Denston has an amazing 60 medieval carved animals. While the great majority are remarkably intact and well-preserved, one – the stag – has had its face chopped right away, leaving only its serated horns to identify it. This appears to have been the case almost everywhere. The stag, mythologically the deadly enemy of the snake (i.e. Satan) seems to have been identified with Christ. Even in bestial form, we may not look on his face.

There is an exquisite crane, identifiable from the bestiary by the stone it carries in its claw, on a misericord in the chancel. The belief was that while cranes slept, one would stand sentinel with a stone in its raised foot: if it fell asleep the stone would fall and wake it up, rather like the dead-man's-handle in a modern railway engine cab. Not all the creatures at Denston, though, are from the bestiary – there are a large number of very life-like rabbits, on the wall-plates and elsewhere, which might almost have hopped in from the surrounding fields.

This magical menagerie is part of what makes Denston perhaps the greatest treasurehouse in all Suffolk for a medievalist, but there are plenty of beasts both lifelike and fantastical to be found elsewhere.

A visit to the area just north of Bury St Edmunds is as good as a trip to a medieval zoo. A number of village churches in close proximity there – Norton, Stowlangtoft, Tostock, Ixworth Thorpe and, not far away, Woolpit – are remarkable for both the quantity and quality of the carvings on the elbow-rests of their medieval benches. Many of them seem to be the work of one prolific 15th century artisan, or perhaps one busy workshop.

Enter the grand 14th century church of St George, Stowlangtoft, by the priest's door in the chancel, as you must outside service times, and you will see at once that the area to your right – the sanctuary – is a no-go area. Not just for hieratical reasons, but because alarms will ring here and at the police station if you touch or cross the altar railing. What they are guarding is an extraordinary range of 16th century biblical carved panels that adorn the east wall. They are known, in the kind of circles where these things are known, as The Stowlangtoft Carvings – which is rather bizarre as they only came here in the 1870s. Their origin is Flemish. DP Mortlock, author of the essential Popular Guides to Suffolk Churches, notes that they were "wickedly stolen" in 1977. He adds: "Although recovered in Amsterdam, by 1986 the parish had still not been able to get them back... Let us hope that persistence will prevail".

Well, it seems persistence has, but I am not sure how glad we should be. The cost of their protection and insurance must be substantial, as no doubt was the cost of their "recovery": all that money might be better spent on preserving what truly belongs here, not something that belongs in Flanders, and might reasonably be "recovered" by the Belgians.

For here, on the common, accessible side of the altar rails are the real Stowlangtoft carvings. They may not be high art in the sense that a sale-room, collector or thief might recognise it (and thank heaven for that), but in real terms they are of inestimable value. And they are as local as the hillock the church stands on.

In 15th century Suffolk, when church-building was a major industry, the carving of elaborate elbow-rests was a flourishing craft - not that you would want to rest your elbow on most of them - and Stowlangtoft was one of the places it flourished most. More than 60 wonderful carvings survive in the nave. Together with the choir-stalls and misericords in the chancel they are as fine a collection as you will find in any parish church.

Though a few of the subjects are familiar from other local churches, the style of carving is not, and many are unique. In its pose, horn scratching its hindquarters, the unicorn may well have been influenced by those at Norton, Tostock and Honington – or perhaps vice versa – but the carving is clearly by a different hand. At Honington and Woolpit there are nice carvings of dogs with geese in their mouths – here the hunting creature is a fox, known from the bestiary as "a clever, cheating animal. the symbol of the Devil". One might reasonably assume that to a medieval countryman, as to a more recent one, a goose caught by a dog was happy hunting, while a bird in the maw of a fox was the Devil's work.

Bench carvings can often be ascribed to such almost folkloric sources; elsewhere there is a clear religious or sacramental meaning. At Stowlangtoft there is a strong sense that the carver just did what he fancied, copying some images from various sources but making others up more or less as he went along. Their apparent absence of saintly, iconic or liturgical meaning is probably what enabled them to survive so gloriously intact through the centuries of reforming and puritan vandalistic zeal.

But are all the carvings in fact so innocent of iconic content? Many of the creatures here are animals with human heads. Why? There seems to be no immediately retrievable answer to that. But perhaps there is.

Mystery creatures - or the four evangelists?

The figure of an old man, cross-legged, writing with a quill on a scroll, an ink-pot in his left hand, is identified by both HM Cautley and Mortlock as Scandal – but on what evidence? Nearby sits what appears to be a bull with a man's head. Could this be intended to represent St Luke, whose evangelistic symbol is the bull? If so, this raises the intriguing thought that what others have taken for a manticore (man-headed, lion-bodied) could stand for St Mark. The man-headed eagle could be meant for St John – and what has been identified as "Scandal" may in fact be an image of the fourth evangelist, St Matthew.

Identifying the mermaid is not fraught with any of this guesswork or detection, but you might not expect to see her with the face of an old woman. In fact almost all the faces here are old. They are all exquisitely carved, all highly individual and with such depth of expression one cannot escape the suspicion that they are portraits of real Stowlangtoft people.

One of the oddest is a creature Mortlock calls a cock monster –"a cock with a hideous human face". But it is not hideous – it is an elderly, bearded face, strong, dignified, full of character. Perhaps what Mortlock takes for feathers was in fact meant by the carver as foliage, making the image not a monster but a green man, that enigmatic creature of the woods whose image, possibly pre-Christian in origin, is familiar from inn signs, carvings on churches, and even the beams of private houses throughout the country. We can never know for sure.

Another strange creature is identified by both Cautley and Mortlock as a cockatrice – a mythological creature said to have been hatched by a serpent from a cock's egg. Like the ancient Greek gorgon or the basilisk, with which it is often associated, the cockatrice was thought to kill with a mere look. A fearsome creature, then, and one often to be found in churches despite its distinctly non-Christian pedigree. The Stowlangtoft creature has much in common with the cockatrice at Tostock, but is that really a cock's head? I don't think so. It looks to me as if the craftsman copied part of something he saw elsewhere, but just decided to do the head differently. It might equally well be a wyvern, a fabulous beast with wings like a dragon, two legs, and a long, curling tail. The cockatrice, identified by the bestiary writers with the Devil, was certainly popular with Suffolk carvers. One of the best is at Stonham Aspall.

Also among the Stowlangtoft carvings is a lovely one of an animal – a rat or vole, perhaps – dressed as a friar preaching. This is clearly satirical, and is again a variation on a theme to be found elsewhere, the apes in friars' habits in wood at Woolpit and in stone on the roof parapet at Bramford.

There is a boar depicted as realistically as you could wish for – except that it is playing a harp. Is there a satirical message in there somewhere too, or is it simply an exercise in the absurd? It has been suggested that the boar has connotations of pre-Christian protective magic, but this seems too great a stretch for the work of a 15th century craftsman who can be assumed not to have considered there might be any other religion than Christianity (and a Roman Christianity at that).

There is a delightful dog that might well have been a portrait of the artist's own companion. In any case, it was undoubtedly carved from the life as, probably, was the rather melancholy chained bear – bear-baiting and dancing were popular in medieval England. The camel, though an equally naturalistic carving, surely cannot be a work of direct observation. There are more camels, more or less recognisable, at Ufford and Tostock.

At Woolpit is another beautiful collection of carved animals – and a trap for the unwary medievalist. For here is the earliest major commission by a known and prolific Suffolk artist, whose lively carvings can be found in no fewer than 83 churches in the county. Originally from Lincolnshire, his name was Henry Ringham, and he perfected the craft of copying, aping – or faking – medieval wood-carving. He died, aged 60, in 1866, by which time he probably knew more than anyone then, before or since about Suffolk roofs and bench-ends. Often it can be difficult to distinguish his work from genuine originals (if, of course, you feel the need to do so – Ringham's work is just as delightful as that of the unknown artists of four centuries before, whose skills and vision he revived). There are clues, however, to help distinguish Victorian work from the true medieval.

First, and most obviously, there is the condition of the carvings. Many that were not destroyed utterly were badly damaged or defaced, either in the Protestant fervour of Reformation in the reign of Edward VI (1547-53) or in the revived fury of intolerance that broke out in the time of Oliver Cromwell, 100 years later. Even if they have escaped the axe, sword and chisel of iconoclasm, elbow-rests that have been rubbed, held and leant on by 20-odd generations of churchgoers are likely to show the effects more than more recent work.

Sometimes Victorian or 20th century craftsmen have repaired damaged carvings: however carefully the new wood has been spliced in there is almost always a tell-tale difference of colour and sheen. A perfect example of this is at Stonham Aspall, where there is a lovely carving of a wolf guarding St Edmund's decapitated head. Close inspection reveals that the wolf's head, though charming, is newer work.

In such cases, older wood is generally darker than newer, often polished by time to a much deeper shine. Where there is no splice or join to be found, a good guess at the age of a carving might be made by an assessment of the bench it is part of: with a practised eye, one can usually tell medieval from Victorian or modern wood, especially if there are examples of both in the same church. (This is not infallible, though: you may find "experts" disagreeing by several hundred years about the age of certain items.)

A more reliable clue might be the presence of a candle-prick hole in the top of an elbow-rest. These holes, perhaps half a centimetre across, were probably made between the late 17th and early 19th centuries, when candles were used simply to light the church – not for religious or ceremonial purposes, as in earlier times. Sometimes there remains a tell-tale trace of molten wax, or simply a wax-polished appearance around the hole. Often the candle-holes have clearly been made after more severe, iconoclastic, damage to the elbow-rest: sometimes they have been gouged into otherwise intact carvings. In either case, their presence is evidence that the carving itself is medieval. With a very small number of possibly 17th-century exceptions, elbow-rests were not carved for about 400 years after the Reformation, and candle-pricks are not generally found in Victorian or later work, since by then oil or gas lamps were in use for lighting.

Often it is possible to compare directly a worn original carving with a later copy in the same church: this is true at Woolpit, where Ringham copied directly from benches still to be seen elsewhere in the church. In fact, much of his work here is in the roof, where 200 wooden angels adorn the 14th-century false-double-hammerbeam structure – all but one of them carved by Ringham in 1844 to replace originals destroyed in the 17th century. His bench carvings seem to consist entirely of good copies of even better late-15th century examples that still survive.

There is a pair of ibex and a pair of griffins. There is a cat which appears to have caught a rat. Also apparently carved from the life are a matching pair of apes, one chained, the other, mentioned above, in the habit of a monk. This one Ringham did not choose to replicate. The joke is clearly at the expense of the itinerant friars who roamed from village to village, preaching as they went, often to the irritation of the incumbent priests. Casting the friar as an ape, a symbol of vanity, indecorum and fraud, is a barbed, almost Chaucerian, comment on a figure expected to preach humility and repentance. Seen purely as art, though, perhaps the best of the Woolpit figures are a number of beautifully observed dogs, including one with a rabbit in its mouth and one with a goose.

Another dog twists its neck back to turn and lick itself – an image naturally observed from life, of course, but also reminding the medieval congregation of a familiar moral from the bestiary: "A dog's tongue will heal a wound if he licks it. His way of life is temperate." The quotation is from Richard Barber's modern translation of the 13th century so-called Oxford Bestiary, now in the Bodleian Library, which also has an illustration of dogs licking their wounds. Similar carvings can be found in a number of churches, notably Lakenheath and Wordwell in Suffolk .

The same gesture of a neck twisted to enable an animal to lick its own flank occasionally occurs in carvings of a more fabulous creature – the unicorn.

"The unicorn... is a little beast, not unlike a young goat, and extraordinarily swift... Our Lord Jesus Christ is the spiritual unicorn of whom it is said 'My beloved is like the son of the unicorns'... The unicorn often fights elephants; it wounds them in the stomach and kills them." This, again, is from the Oxford Bestiary. There is also a medieval tale of the unicorn being trapped by its habit of sleeping in a virgin's lap. All this would have been well known by the superb craftsmen of the early 15th century who carved unicorns on both the font and arm-rests in St Andrew's, Norton.

The superb bench carvings here include possibly two, or even three, unicorns, though all – perhaps not surprisingly – have lost their horns. There are two creatures licking themselves – the movement of neck and tongue beautifully realised, even if the actual beasts cannot be identified with certainty. There is a cockerel, again related to the cockatrice in neighbouring Tostock – this time so closely related one can be pretty sure they are the work of the same chisel-hand. And there is a stag, identifiable by its one surviving serated antler even though, as at Denston, it has been savagely defaced, losing most of its head.

Four creatures support the shaft of the Norton font, including a mysterious big-chinned fellow of bizarre and demonic appearance, a lion, and an animal that may be a sheep or a goat but looks precious little like any the carver ever saw. If it is a goat, it may be related to the goat badge of the de Bardwells, who were lords of the manor at Norton at that time and whose shield appears twice in a north aisle window. (The rampant goat can also be seen at the grand church of the village of Bardwell itself, in the spandrels of the porch.)

The fourth creature is a wodewose, or wild man of the woods, possibly the commonest carved figure to be seen in Suffolk churches. For more on the wodewose, see the chapter The Mystery of the Wodewose.

There are more splendid survivals among the bench carvings at Dennington, two of them apparently unique in England. One is supposedly a tortoise, though you will need some imagination to identify it with much certainty. The other, probably the best-known attraction of a church that demands to be visited for many reasons, is the sciapod. The known pedigree of this fabulous fellow goes back to Pliny, a second-century Roman provincial governor whose letters to the emperor Trajan provided the medieval bestiary-writers with one of their favourite sources. He describes the sciapod as having one enormous foot under which he lies shaded from the sun. A renowned great leaper, the sciapod has a hop-on part in Voyage of the Dawn Treader, one of CS Lewis's much-loved Narnian chronicles, where he goes by the name of Dufflepud. His Dennington incarnation appears to have been copied from an illustration in the 13th century Westminster bestiary, or one like it, though the carver has given him two legs and feet.

Here also, among the elbow-rests, are a very alert-looking dog, a lion with a gloriously feathery mane, a mermaid – younger-looking but less well-preserved than her sister at Stowlangtoft – and an animal generally identified as a wolf, though I am not convinced (it appears to have cloven hooves). There is a two-headed eagle, which may just have a common eastern ancestry with the one taken about the same time as a symbol by the Russian Tsar Ivan III. And there is another of those beasts, like the Denston elephant, which appears to have been carved from a verbal description. Horse-like from the shoulders down, it has a long, spotted neck which identify it as a giraffe – though the neck is curled down and back in a manner you would not see on the African plains. And there is a rather fine, though headless, example here of probably the best-known bestiary image, the pelican in her piety.

Bearing scant resemblance to the bird we now call by that name, the pelican's foremost attribute was restoring her dead chicks to life by feeding them from the blood of her own breast. In this way she was likened to Christ sacrificing himself for the redemption of his "children", mankind.

Images of the pious pelican abound, most of them dating from the 15th century: one striking example is at the pinnacle of the towering font-cover at Ufford, and another on a corbel supporting the roof at Stoke-by-Nayland. Good pelican elbow-rests can be found at Tuddenham St Martin, Great Bealings, Combs, Earl Stonham and Great Waldingfield; there are pelicans on the fonts at Risby and Pakenham, and one in a fragment of medieval glass at Pettistree.

It is quite plain, when you consider the medieval bestiary today, that the writers and artists made no distinction between real beasts and mythological ones. The Denston elephant is just as fantastical as the Norton unicorns; the Dennington mermaid scarcely stranger than the giraffe. So how, if you were a late medieval churchgoer, would you identify a tiger? Why, by its mirror, of course.

The tiger, almost a sermon in itself, was certainly used – like all the other bestiary creatures – to illustrate medieval sermons. Indeed, it was from sermons that most people would have had their only knowledge of the fabulous animal. The tale was that a man riding off with a tiger cub would be certain to be caught by the enraged mother unless he confused it by throwing down a mirror or a glass sphere - the tiger, seeing her own reflection in the glass, would take it for her young, and so allow the thief to escape. "So the intensity of her motherly love betrays her and deprives her of both her revenge and her cub." It may he hard to imagine what useful moral might be derived from this legend, but if one compares the painted tiger in the Oxford bestiary with the carved one on a bench-end at Lakenheath, it is clear to see where the carver took his original. The composition is essentially the same, although the differences – the carved tiger's mane and hooves and the long handle on the mirror – are enough to suggest he had a similar but slightly different book to work from.

All the mermaids we have seen also hold mirrors. This is not to ward off tigers, but again to illustrate common sermons. One association is simply the supposed vanity of the female, and links the mermaid, who might lure sailors to their death, with the dangerous lure of the prostitute. Another, perhaps more challenging for the preacher, suggests that the mirror reflects reality in the same way that the physical world reflects the divine. In this way, the carvings themselves mimic both the real and imagined worlds, making wonders solid in the daily reality of the Suffolk village.