Saturday, June 03, 2017

What is myth?

(This was originally prepared in 2013 for students taking the Open University module A330 Myth in the Greek and Roman Worlds. It represents my personal opinions, not those of the  OU Module Team, and is not part of the official module materials. I've now moved it over here because I'm leaving the OU, and want to continue using this in my teaching. I've made some minor amendments.)

The question I want to address briefly here is ‘what is myth?’ This is, of course, a pretty fundamental question for most courses on mythology, including the Open University module A330 Myth in the Greek and Roman Worlds, and it does get addressed in the A330 Course Introduction (Emlyn-Jones and James, 2011, pp. 8–10). But after that, students on that course, or indeed on many other courses on myth, may find that the question gets slightly forgotten. This is because, in a lot of ways, it’s really quite a hard question. And the reason for this is that myth is one of those things where everybody knows what it is — but when it comes to seeking a hard-and-fast definition, that turns out to be a lot more difficult than you might expect.

As the Open University Course Introduction says, there are lots of different meanings for the word ‘myth’ (Emlyn-Jones and James, 2011, pp. 8–9). If you own a copy of the Open University publication The Arts Good Study Guide (Chambers and Northedge, 2008; a text, incidentally, which I recommend to anyone studying an Arts or Humanities course), you will find in there a section on ‘Some myths about exams’ (pp. 285–290). Of course, the authors are not talking there about a heroic struggle of a legendary figure to pass the examination for a course in mythology. They are instead using ‘myth’ to describe something that is commonly believed about exams, but is not (in the opinion of the authors) actually true. But this is not what we mean when we speak of ‘the myth of Heracles’ – because no-one actually believes that the story of Heracles is literally true.

The word ‘myth’ comes, of course, from the ancient Greek word μῡθος or muthos (sometimes transliterated as mythos). As students on the OU module Myth in the Greek and Roman Worlds learn in Block 4 (Emlyn-Jones in James et al., 2011, pp. 112–113, 189–192), this originally simply meant ‘the spoken word’, and was contrasted with logos, the written word. Gradually, the meanings of both terms shifted, so that from its basic meaning, muthos came to mean a spoken story, such as, for instance, Homer’s Odyssey, recited by a rhapsode. That led to muthos having connotations of the made-up, and hence the irrational and the false, whilst logos in consequence represented the rational and the true. And as part of that, muthos came to represent, for the Greeks, what we would now describe as ‘mythology’.

But for a long time, ‘myth’ wasn’t the term that post-Classical scholars used. They preferred the Latin term fabulae, which means ‘stories’, or ‘tales’ (and later became the standard Latin term for a dramatic performance). The Greek term muthos only came into favour with the work of the German scholar Christian Gottlob Heyne (1729-1812), who found fabulae too slight for the weight he wanted to place on mythology, and so started using the Greek term muthos (also, as Morales, 2007, p. 57, notes, he wanted to use the contrast between muthos and logos to advance his case that the Greek myths belonged to a primitive stage of humanity, as opposed to the rationality of the eighteenth-century Enlightenment). From there muthos passed into English-speaking scholarship as ‘myth’.

But still, what is a ‘myth’? Mark Morford and Robert Lenardon, in their textbook Classical Mythology, do advance a definition in their first chapter. They say:

classical myth is a story that, through its classical form, has attained a kind of immortality because its inherent archetypal beauty, profundity, and power have inspired rewarding renewal and transformation by successive generations.

(Morford et al., 2015, p. 26)[1]

Is that definition acceptable? Well, for a start, it is, as Morford and Lenardon state, only a definition of a classical myth; it won’t do as an overall definition of ‘myth’, which has to encompass all the other mythologies, such as Norse, Chinese, Sumerian, etc. But even if we removed the specific reference to the Greek and Roman world, will what Morford and Lenardon say work as a definition of myth? I think problems remain. For one thing, this definition really only covers the famous and often retold myths. There are, however, many stories that are obscure, and rarely retold, such as Ovid’s tale of the crow in Metamorphoses, Book 2 (549–595). Are we to remove those stories from mythology, because they haven’t ‘inspired rewarding renewal and transformation’? I think not.

Another course on Greek and Roman mythology, run as an online MOOC by Peter Struck of the University of Pennsylvania for Coursera, adopted the following definition of myth:

a traditional tale with secondary partial reference to something of collective importance… told by someone for some reason.

(Struck, 2012a)

Struck is drawing his definition from that offered by Walter Burkert in Structure and History in Greek Mythology and Ritual:

myth is a traditional tale with secondary, partial reference to something of collective importance.

(Burkert, 1979, p. 23)

Burkert's definition has been very influential; but it is a pretty vague sort of definition, one which covers a lot of ground. It’s certainly better than the heavily value-weighted definition of Morford and Lenardon. But even Struck's course had to throw out this wishy-washy definition when it came to looking at Ovid’s Metamorphoses, a work OU students study in some depth in Myth in the Greek and Roman Worlds. Struck replaced his original definition with this definition of mythology:

a clump of stories, probably untrue, about ancient characters that may or may not have actually existed, with some deeper truths in them, or not. But they’re surely fun to hear.

(Struck, 2012b)

The problem is, of course, that any mythology includes a whole range of different stories, created for different purposes. There are aetiological myths, myths that explain the origins of a feature of the natural world. This sort of story used to be familiar to children through Rudyard Kipling’s Just So Stories (1902), which explain various features of animals, such as ‘How the Leopard got its spots’, and ‘How the Elephant got its trunk’. Ovid’s Metamorphoses is full of such stories, as anyone who reads that work will find out — one example from Book 2 (401–531) is the myth of Callisto, explaining the origins of the constellation of the Little Bear, and why that constellation never sinks below the horizon (from a Mediterranean perspective).

Then there are myths that probably have their origins in actual historical events, such as those myths concerning the Kings of Rome (though too much assumption of a historical background behind a myth such as that of the Trojan War brings its own problems).

And there are myths that are just funny stories, with no particular object beyond that. An example of this might be the story of Iphis and Ianthe in Metamorphoses Book 9 (666–797), a humorous tale about a girl raised as a boy.

And this is just to cover a few types of myth.

In their first chapter, Morford and Lenardon (2015, pp. 3–39)[2] advance a possible method of rationalizing this complexity. They address the manner in which some scholars have divided mythology into ‘true myth’, which concerns the gods, ‘legends’ or ‘sagas’, which have a historical basis, and ‘folktales’, those stories intended as pure entertainment. They add a couple of subcategories of folktales: ‘fairy-tales’, those folktales with a high moral and/or magical content; and ‘fables’, for stories in which animals are the principal characters, best-known from the tales of Aesop, such as ‘The Hare and the Tortoise’. They make no distinction between ‘legends’ and ‘sagas’, though one could argue that ‘legends’ should be individual stories, and ‘sagas’ longer linked stories around a central character or event (such as the Trojan War).

There are several problems with this rationalization. One is the unnecessary importation of the term ‘saga’, a specific term from Norse literature, into the study of Greek mythology. Another is that not everyone uses terms in the same way — for example, J.R.R. Tolkien, in his 1939 essay ‘On Fairy-Stories’ (in Tolkien 2012), clearly includes in the category of fairy-stories the sort of high fantasy/mythology he had been writing since the 1910s, and which would soon be published in The Lord of the Rings.

But the biggest problem for this rationalization is the one that Morford and Lenardon themselves identify (2015, p. 4)[3] — that few actual myths can be fitted neatly into these categories, and most include different elements taken from different categories.

This is a natural product of the constant retelling of the stories. A story intended to explain some element of the world around the original storyteller will then be retold for a contemporary purpose by an Athenian tragedian, or as an entertainment by Ovid. A mythology is built out of ‘true’ myths, legends, folktales, fairy-tales and beast fables, so it is important to know that some people make these distinctions—a former colleague of mine amongst the Open University Associate Lecturers, for instance, distinguished between the story of Aeneas, a myth of the foundation of Rome, and the story of Romulus, a legend of the foundation of Rome. But trying to categorize all mythological stories is fraught with problems, and I will continue using the general term ‘myth’—and should I use other terms such as ‘legends’, I don’t mean anything specific by this.

I'm still no closer to a precise definition of myth. But perhaps I shouldn’t be looking for one. In fact, there are a lot of concepts that resist easy definition in this way. As part of my scholarly portfolio, I do some work as a science fiction critic. Science fiction critics get very exercised about how to define science fiction. One famous definition was offered by the critic Damon Knight in 1956, in a work called In Search of Wonder. He says:

… it will do us no harm if we remember that [science fiction] means what we point to when we say it.

(Knight, 1996, p. 11)

That sounds like an enormous cop-out—it doesn’t try to tell you what science fiction is, it just tells you that you’ll know it when you see it. But in fact it’s the only definition that makes any sense. Most of us have a fair idea of what the core of science fiction is, but we’ll always be arguing about works on the edges of the genre. And those arguments will always undermine any attempt to more precisely define what we mean by ‘science fiction’.

The same thing happens in other genres. Let’s look at western movies, for instance. Everyone knows that The Searchers (USA, John Ford, 1956) or Butch Cassidy and the Sundance Kid (USA, George Roy Hill, 1969) are westerns. But if one adopts a definition that says, for example, that westerns are set on the American frontier between the American Civil War and the end of the nineteenth century, one ends up excluding movies such as Red River (USA, Howard Hawks, 1948), which begins in 1851, a decade before the Civil War. Even if one stretches the chronological boundaries, to, say, from the Texas Revolution of 1835 to the start of the First World War in 1914 (or US participation in 1917), one excludes Bad Day at Black Rock (USA, John Sturges, 1955), set in the late 1940s, yet often included in lists of the best westerns ever.

To borrow a term from semiotics, science fiction, westerns, and mythology are all ‘analogical’ modes of communication, where meaning is articulated through proportion, expression and gradation, as opposed to ‘digital’ modes, where clear-cut definitions are employed. This means that there is general consensus about the core of such concepts, but there are grey areas at the edges, where interesting discussions can be had. As a mythological example of a grey area, one can consider the degree to which the emperor Augustus not only employs mythological imagery, but becomes himself, through the writings of Virgil and others, a mythological character. The same thing happens to Julius Caesar at the end of Ovid’s Metamorphoses (15.745–870). Can one therefore talk about the ‘myth of Augustus’ or the ‘myth of Caesar’? Obviously people do, but generally they mean the image that Caesar and Augustus presented of themselves, not Caesar or Augustus as mythological characters with the same status as Odysseus or Heracles.

So I can’t provide a definition of myth. But you may have noticed that one word has recurred through these discussions. That word is ‘story’, and I would certainly assert that the principal feature of myths is that they are stories.

If no single definition of myth is universally applicable, it follows, as Morford and Lenardon say (2015, p. 3),[4] that no single theory of myth explains the entirety of myth. This hasn’t, of course, stopped proponents of various theories asserting that their theory explains everything, whether that be Bronislaw Malinowski and the functionalists, who argue that myths legitimized aspects of society, or the Freudians and their idea that myth represents the collective unconscious of humanity. Many of these are ideas that can be useful for understanding individual myths — well, perhaps not Freud, but that’s an argument for another time (for a partial takedown of Freud, see Morales, 2007, pp. 74–79) — but they don’t explain everything. Familiarity with theories of myth is an important part of any course on myth. But inevitably, those theories must be used with caution.

I’ll end by saying something about what myth is not. Myth is not something that exists independently of the texts, literary or visual, in which it is recorded. This may seem obvious — of course we can’t experience myth except through reading or looking at images. But it is surprising how much the idea persists that the representations of myth are reflections of some concept that is referred to as the ‘original myth’, rather as Plato theorized that everything we experience is a manifestation of an ideal concept.

However, there is no ‘original myth’, at least not in any sense that is meaningful to us. Of course, there will have been a time back in the mists of prehistory when the story of Odysseus was first told. But we cannot recover that moment, or the form that the story took at that telling. We can talk about the earliest versions recorded, but even those are the products of countless earlier retellings that we no longer have access to.

We can also talk about ‘dominant’ versions. These are the versions that become the ones that people first think about when they consider a myth, and which later versions engage with and react against. These are not necessarily the same as the earliest surviving versions. For the story of Odysseus, Homer’s Odyssey, the earliest surviving version, is also the dominant version. But for the story of Oedipus, the dominant version is in Sophocles’ play Oedipus Rex, and this differs significantly from the earliest surviving version, which is again that of Homer’s Odyssey. Sometimes these dominant versions are called ‘canonical’, but that implies that these versions are correct, and versions that differ are wrong. This is not, in my view, an idea that is useful.

Because there is no one ‘right’ version of a myth, it is impossible to treat sources for myths in the same way as one might treat historical sources. Many students trained on historical courses will be used to taking various sources, and fitting them together like a jigsaw, to create as true a picture as possible of the event. But in myth, that simply can’t be done. We have a collection of jigsaw pieces, yes, but each piece comes from a slightly different jigsaw. If those pieces are forced together, the picture created is inevitably a distorted one.

So this leads to a final comment: what do you mean when you talk about, for example, ‘the myth of Hippolytus’? Do you mean an embracing concept that includes all the variations on that myth? Or do you mean the version in Euripides’ play Hippolytus, which is the dominant version? Or do you mean a composite version that has been put together for the Penguin Dictionary of Classical Mythology (Grimal, 1991, p. 204), or some other reference work? You’ll have effectively grasped hold of the essential concepts of courses on myth when you are able to mean the first, rather than the other two.

Reference list

Burkert, W. (1979) Structure and History in Greek Mythology and Ritual, Oakland, CA, University of California Press.

Chambers, E. and Northedge, A. (2008) The Arts Good Study Guide, 2nd edn, Milton Keynes, The Open University.

Emlyn-Jones, C. and James, P. (2011) A330 Myth in the Greek and Roman Worlds: Course Introduction, 2nd edn, Milton Keynes, The Open University.

Grimal, P. (1991) The Penguin Dictionary of Classical Mythology (trans. A. Maxwell-Hyslop; ed. S. Kershaw), London, Penguin Books.

James, P., Hughes, J. and Emlyn-Jones, C. (2011) A330 Myth in the Greek and Roman Worlds. Block 3: Ovid and the Reception of Myth; Block 4: Myth and Reason in Classical Greece, 2nd edn, Milton Keynes, The Open University.

Kipling, R. (1902) Just So Stories, London, Macmillan & Co.

Knight, D. (1996), In Search of Wonder: Essays on Modern Science Fiction, 3rd edn, Chicago, Advent.

Morales, H. (2007) Classical Mythology: A Very Short Introduction, Oxford, Oxford University Press.

Morford, M.P.O. and Lenardon, R.J. (2007) Classical Mythology, 8th edn, Oxford & New York, Oxford University Press.

Morford, M.P.O., Lenardon, R.J. and Sham, M. (2011) Classical Mythology, international 9th edn, Oxford & New York, Oxford University Press.

Morford, M.P.O., Lenardon, R.J. and Sham, M. (2015) Classical Mythology, international 10th edn, Oxford & New York, Oxford University Press.

Struck, P. (2012a) ‘Week 1 Lecture 4: Ideas on myth from the modern era’, Greek and Roman Mythology [Online]. Available at (Accessed 15 September 2014).

Struck, P. (2012b) ‘Week 10 Lecture 4: Ovid—Background and themes’, Greek and Roman Mythology [Online]. Available at (Accessed 15 September 2014).

Tolkien, J.R.R. (2012) The Monsters and the Critics and Other Essays [ebook reader], London, HarperCollins.

[1]      In the 9th edn (Morford et al., 2011) this text can be found at p. 25, and in the 8th edn (Morford and Lenardon, 2007) at p. 29.

[2]       8th edn (2007), pp. 3–39; 9th edn (2011), pp. 3–36.

[3]       8th edn (2007), p. 4; 9th edn (2011), p. 4.

[4]       8th edn (2007), p. 3; 9th edn (2011), p. 3.

Saturday, May 06, 2017

2017 SFF Masterclass - places still available

Please share.


Three days of extremely enjoyable discussion and exchange of ideas and in the delightful environment of the Royal Observatory Greenwich, the Masterclass is highly valued by past students. Places are still available on a first come-first served basis. Applications welcomed from Past Masterclasses students.

To apply please send a short (no more than 3,000 words) piece of critical writing (a blog entry, review, essay, or other piece), and a one page curriculum vitae, to

For more details, see:

Saturday, April 01, 2017

2017 Science Fiction Foundation Masterclass in Science Fiction Criticism - Applications open

Applications are now open for the 2017 Science Fiction Foundation Masterclass in Science Fiction Criticism. The 2017 Masterclass, the Eleventh, will take place from Friday 30 June to Sunday 2 July. We are delighted to have once again secured at the Royal Observatory Greenwich as a venue.
The 2017 Class Leaders are:
Nick Hubble (Brunel University)
John J. Johnston (Egypt Exploration Society)
Stephanie Saulter (author of Gemsigns and sequels)
Price: £225; £175 for registered postgraduate students.
To apply please send a short (no more than 3,000 words) piece of critical writing (a blog entry, review, essay, or other piece), and a one page curriculum vitae, to
Applications received by 24 April 2017 will be considered by an Applications Committee. Applications received after 24 April may be considered if places are still available, on a strictly first-come first served basis.
Past Masterclass students are encouraged to apply again (though we will prioritise applications from those who have not been previous students).
Information on past Masterclasses can be found at Please direct any enquiries to
Tony Keen, SFF Masterclass Administrator

Sunday, February 19, 2017

A National Gallery tour

[This is a post I originally put on my OU blog on 25 August 2010. I'm reposting it here so non-OU students and colleagues of mine can see it (back then I didn't have such things). Obviously, some of the paintings not on display then now are, and vice versa. And these days my route would be slightly different, including Bronzino's An Allegory with Venus and Cupid, and Lastmann's Juno discovering Jupiter with Io.]
At the beginning of this month I took my last group of students round the National Gallery in London as a post-exam treat. I've been doing this for a long time, and I hope that students appreciate it. The intention was to have a look at some paintings related to the Classical world, with one eye on A330 Myth in the Greek and Roman Worlds. [These were students who had done the module on Exploring the Classical World, and might be tempted to the follow-on module.]
Anyway, the tour went as follows:
I began chronologically at the end, with Room 34: Great Britain 1750-1850. I began here simply because it seemed to make most sense in terms of the Gallery's layout, from the main entrance. I had come here for J.M.W. Turner's Ulysses Deriding Polyphemus - Homer's Odyssey. Turner didn't do a lot of Classical landscapes, but this one always appeals to me. I'm never quite sure if the Cyclops can be seen in this painting (is he the dark shape above the ship? or are those just rocks?), but there's lots of details you can point out, such as the almost-invisible horses of the sun, or the Trojan War depicted on the ship's flag (I nearly got in trouble from the attendants for getting too close when pointing that out). It's nicely placed next to The Fighting Temeraire.
We passed swiftly through Room 33: France 1700-1800, pausing only briefly for Psyche showing her Sisters her Gifts from Cupid. We paused only slightly longer in Room 32: Italy. There we looked at Dido receiving Aeneas and Cupid disguised as Ascanius and, on the opposite wall, Perseus turning Phineas and his Followers to Stone. With regard to the latter, I talked about how many of the mythological scenes depicted in post-Renaissance painting are derived from Ovid's Metamorphoses. In that context, I should really have discussed The Fall of Phaeton, but I didn't.
In Room 30: Spain, we stopped at The Toilet of Venus. Here we talked about how, in this commission for the Marqués del Carpio, Diego Velázquez was able to use the cloak of making the painting mythological essentially to allow him to depict a real naked woman, in a manner that would otherwise be unacceptable - if not for the addition of wings to the boy, this would have been considered wholly immoral. Even then, it was probably not displayed publicly. And it has remained controversial, as shown when a suffragette slashed it in 1914. I also find it interesting that when Manet, to all intents and purposes, showed us the other side of this image in Olympia, whilst he stripped away much of the mythological associations, he left her with a Classical name.
Room 29: Peter Paul Rubens, has paintings that we could have looked at, including two versions of The Judgement of Paris. Instead, we headed for Room 12: Titian and Venice 1500-1530 [this has now moved to Room 2]. The key piece here is Bacchus and Ariadne. I really liked talking about this painting, because there is so much one can get out of it. It crystallizes the moment that Ariadne ceases to care about Theseus, who has abandoned her (Theseus was, frankly, a bit of a [four letter word], at least in this instance). This is symbolized by making his ship very small, and placing it on the opposite side of Ariadne from where she is looking. But there is so much more: Ariadne's crown set in the sky as stars, the chariot of Bacchus in the form of a sarcophagus, thus symbolizing resurrection, the sly nod to the Sistine chapel in Bacchus' arm. (I'm grateful to Chris Wilson for pointing that out in a previous National Gallery trip that he led, and which he is repeating for the London Region Arts Club on Sunday October 17.)
We also went across the intersection to Room 10: Venice 1530–1600 [now Room 6], for Titian's unfinished The Death of Actaeon. Here, once again, one can see the influence of Ovid's Metamorphoses. Though the transformation into a stag is in the earliest versions of the myth, there are other versions (in vase painting, etc.) in which the transformation appears not to take place, and it is Ovid's terrifying description in Metamorphoses 3.138-255 of Actaeon being torn apart by his own hounds, unable to communicate with them, has become the dominant version. Unfortunately, Titian's companion piece to this painting, Diana and Actaeon, is not currently on display - it's jointly owned with the National Galleries of Scotland, and spends five years at each institution, so at the moment it's in Edinburgh. [It is now, of course, in London.] (Charles Martindale has an interesting discussion of Titian and Ovid, including these paintings, in Redeeming the Text, pp. 60-64.)
From there it was off to Room 19: Nicholas Poussin, principally for two treatments of essentially the same subject, A Bacchanalian Revel before a Term of 1632-1633, and The Triumph of Pan, from 1636. We compared the two paintings, and also compared them with the same artist's The Adoration of the Golden Calf, and Titian's Bacchus and Ariadne. I feel that the revel in the earlier painting looks like mostly harmless fun, whilst that in the later one has the air of an event about to get out of control.
From there, it was into the Sainsbury Wing for the mediaeval paintings. Actually, there was only one room we were interested in, Room 58: Paintings for Florentine Palaces. When we arrived, another tour group was already there, so we listened to their guide talk about A Satyr Mourning Over A Nymph. When they departed, we went to look at Antonio del Pollaiuolo's Apollo and Daphne [the National Gallery now attributes this to his brother, Piero], which depicts the climax of the Daphne myth (scroll down), the point where the nymph Daphne is being transformed into a laurel bush to save her from the god Apollo's amorous intentions. The Penguin translation of the Metamorphoses, which is a set book for A330, has on its cover Bernini's sculpture of the same moment, in which Daphne is being transformed from arms downwards, as in this painting, but also from her legs upwards. The Bernini is probably a better encapsulation of frustrated desire, but it's interesting to compare this to the earlier (by about 150 years) painting (which, incidentally, is used - though for some reason reversed - as the cover for the Penguin edition of Arthur Golding's Elizabethan translation, which Penguin has in print as a classic of English literature).
Unfortunately, the painting that I would have like to conclude with, Botticelli's Venus and Mars, is not currently on display in this room, being shown in the exhibition Close Examination: Fakes, Mistakes and Discoveries, next to a painting that was bought together with Venus and Mars in 1874, and considered at the time to be another Botticelli, though it patently isn't. What I like about Venus and Mars is the contrast between Mars and Venus. The former is asleep, clearly exhausted after a bout of lovemaking. Venus, on the other hand, is wide awake, with a rather inscrutable look on her face. I wonder if she is going through the experience of women through the ages, the post-coital thought, "was that it?"
This, of course, doesn't get close to covering all the paintings in the National Gallery's collection that have mythological subjects, or even all the ones on display (I tried and failed to spot Claude's Landscape with Aeneas at Delos as we passed). The Gallery used to publish a useful Pocket Guide on Myths and Legends. This covers some of the most significant mythological paintings in the collection, including some, though not all, of the ones I mention above. It no longer appears available through their online ship, but you can get it at Amazon, though their stock is almost exhausted. [Now it's via Amazon Marketplace.]

Monday, January 02, 2017

Sherlock, 'The Six Thatchers'

(It's impossible to talk about 'The Six Thatchers' without including spoilers. Hence, there are spoilers here, eventually.)

So now we get the other Steven Moffat series that's been off the screens for exactly a year. How did this one fare? Well, it was probably better than Doctor Who; as the first episode of a new season, it is perforce more substantial than the throwaways that Doctor Who Christmas episodes seem to be now. And it was certainly better than the self-indulgent mess that was last year's 'The Abominable Bride'.

Indeed, there are quite a few moments in 'The Six Thatchers' that remind us of why we became fans of Sherlock in the first place. If Mark Gatiss' script tends to assume that the audience likes the characters, rather than giving them reasons for doing so, it nevertheless has a number of good jokes and quite a lot of clever dialogue. Some scenes definitely had me smiling. There's a neat twist on the original 'Adventure of the Six Napoleons', though one might feel Margaret Thatcher is still perhaps too divisive a character to be used in this role - many will relish the smashing up of her image, others will be offended by it.

Benedict Cumberbatch is very much back on form. Other performances, however, with one exception, are a bit muted. There are some other problems - there's an opening scene, largely designed to get the writers out of a hole they shouldn't have got into in the first place, in which Sherlock appears to have been possessed by the spirit of Peter Capaldi's Twelfth Doctor. And, given the reliance of the show on the John Le Carré international espionage aesthetic for a number of sequences, it's unfortunate, if beyond the showrunners' control, that television audiences have in 2016 been exposed to real Le Carré in the form of The Night Manager, which does this sort of thing so much better. (Incidentally, the 2016 edition of Le Carré's novel includes a fascinating afterword on the experience of seeing his book filmed.)

But overall, 'The Six Thatchers' was doing okay. And then it fell at the final hurdle. And here come the spoilers.

The other first rate performance in this episode is, of course, that of Amanda Abbington. Finally, she gets to convince the viewers that Mary Watson really is a superspy and international woman of mystery. But she's only given this leeway because it's her last opportunity.

That Sherlock would eventually fridge Mary was not exactly unexpected, not least because Arthur Conan Doyle killed off his own Mary. I was quite surprised that she made it through Season 3. The moment that the contents of the last Thatcher turns out to be an AGRA data stick rather than the expected pearl, it's clear that she won't be alive by the end of the episode.

But, as with the death of Osgood in the Doctor Who episode 'Death in Heaven', the creators seemed to feel that they had to kill the character off, but couldn't work out how to do it properly. In this case, Mary's death is pretty contrived. I was expecting the villain Norbury instead of shooting Sherlock to shoot the tank in the aquarium, which would have been a surprising, if unrealistic scene - but this was probably beyond the show's budget. More seriously, Mary seems to need some sort of super-speed to actually get in front of the bullet fired at Sherlock, given that she doesn't respond until it's already left the pistol barrel.

The disappointment here is that there was an opportunity here to subvert genre tropes, but instead the creators chose to lazily follow them. They probably take the view that Sherlock is ultimately about the relationship between men, and women get in the way of that (an inherent drawback of a 130-year old franchise, that can be overcome, but only if creators make an effort). And so we will get at least ninety minutes of John Watson's manpain.

Sherlock, 'The Abominable Bride'

[Another review from the archives, of last New Year's Sherlock.]

When I saw the trailer for 'The Abominable Bride', I wondered if this might actually be Benedict Cumberbatch playing Sir Arthur Conan Doyle's Sherlock Holmes for once. Of course, that doesn't happen - instead we get the usual much less likeable, much more arrogant and often bullying version that we've seen for three seasons of Sherlock.

Nor do Steven Moffat and Mark Gatiss go the whole hog and do an actual Victorian Alternate Universe version of the show - instead, this is all going on in Sherlock's mind, and is therefore of no real consequence. It takes place in the immediate follow-up to the end of Season 3, where Sherlock had committed murder largely because Moffat and Gatiss could find no other way out of the plot corner they had written themselves into, and then used the return of Jim 'Thighs' Moriarty, whom they seem to have turned into a supervillain, as a deus ex machina so that they could continue to make episodes. To be honest, I'd have preferred the AU.

As Dan Hartland observes, in a piece that's well worth reading in full, 'The Abominable Bride' is primarily an exercise in self-referentiality. It's oh-so-clever, down to the final scene that is meant to have you wondering whether it's the 2010s or 1890s Sherlock who is real. This is the same trick that Moffat pulls at the end of the Doctor Who episode 'Last Christmas', where a final tangerine implies that the Doctor and Clara Oswald are still dreaming (which may be a great get-out from the succeeding two years of continuity). The device was frankly old when Buffy the Vampire Slayer did it in 'Normal Again'. These days, I am past caring.

Neither do I care for the overly blokish 'look, we're feminists too' pronouncements with which the episode is littered, about as convincing as David Cameron saying that he really is on the side of the poor (though, as has been pointed out to me, if this is all in Sherlock's head, then the feminism is a priori going to be caricatured). But the biggest problem is that what there is not in this episode is much in the way of story, and what there is is pretty confused.

There are good performances, of course. In particular, Rupert Graves' slightly confused Inspector Lestrade works much more effectively in the Victorian context than in the twenty-first century, where one feels he would be rapidly sidelined from any real work. Martin Freeman as John Watson does well with a set of lines that seem designed to make him look as stupid as possible, the error made by the otherwise admirable Basil Rathbone/Nigel Bruce versions of Holmes and Watson. If this is what Sherlock actually thinks John Watson is like, and the words are not Watson's but Sherlock's imaginings of Watson's, then it's hard to see how their friendship survives.

But these performances are not enough. In short, 'The Abominable Bride' is a mess. A very pretty and professionally made mess, but, as with recent Doctor Who (though it is better than that), arguably evidence that the showrunners are out of ideas, and continuing to do the show merely out of habit.