In 1975, Gary Gygax, wrote lavish praise in the foreword to M.A.R. Barker's Empire of the Petal Throne, calling it the 'most beautifully done fantasy game ever created' and saying Barker's world, Tékumel, was - except for Tolkien - completely without peer. Gygax concludes that the primary difference between the two (Tolkien and Barker) is that the latter "has neither had the opportunity to introduce and familiarise his Tékumel by means of popular works of fiction".1
In 1984, that opportunity came, as Barker's The Man of Gold was published by DAW, the first of five novels set in the universe. Harsan is a young novice in the temple of Thumis, the Lord of Wisdom. His speciality is linguistics, and at the start of the book, he's just about the wrap up his thesis - a study into one of Tékumel's long-dead languages (there are a lot of them - it is an old, old world, built on the ruins of a long of old, old cultures). His bucolic - if dull - monastic existence is interrupted by a messenger sent from the Tsolyani empire.
The Tsolyani are at war with the Kingdom of Yan Kor, and the latter are equipped with an ancient artifact, the tautly-named 'Weapon Without Answer'. Rumor has it, there may be an answer - at least, a crumbling manuscripts says so. But the Tsolyani need someone proficient in the Llyani language to work out the details. For example: Harsan.
Thus beginneth the quest. Harsan travels across the empire to present himself to his superiors - a journey not without its own intrigues (and, um, foxy companions). Once at his destination, he finds himself entangled in a whole big bucket of plots - the various temples, political movements and Imperial heirs all have schemes for the new artifact. And, of course, Harsan. And that's not counting the plotting within each of the factions. Or the very non-human manipulations of the land's very non-human races. And the Yan Kor. And, for that matter, the gods themselves.
Fortunately, although the intricacy is all there - Harsan doesn't partake as much as surf. He's chucked from location to location like so much driftwood. Occasionally this takes place under his own power (in that, he's being openly railroaded into a decision), sometimes it isn't even that subtle, and he's drugged, imprisoned or kidnapped. The locations and scenarios - from the cities to the undercities to the temples to the dungeons - are all entirely fascinating, which is good, since Harsan himself is completely devoid of agency.
As a result, this is a curiously unadventurous adventure. Despite the elaborate (and, frankly, gorgeous) world-building, there's little to care about. Harsan isn't even a compelling blank; he's actually bit dumb, an awkward combination of naive and reckless, traits that are more about fuelling the plot than forming any sort of cohesive personality. Even his romances are oddly charmless. Harsan is constantly pestered by nude or nearly-nude women, which, like the book's weirdly indecisive romantic 'pay-off', seems less about cultural world-building than it is a misguided attempt to provide prurient entertainment. The novel is more escapist when it doesn't try so hard. The setting alone is fuel for the imagination; when it edges into the area of overt fan-service, the writing's fumbling ineptitude becomes painfully clear.
The temptation is, therefore, to declare The Man of Gold a 'bad' book. It lacks compelling characters and the plot is a MacGuffin-fueled shambles. The world - the setting - is glorious, but as a novel... hell, as a story, this is pretty rubbish.2 But The Man of Gold is actually a resounding success. Skip back to Gygax - another [very] 'bad' novelist - and think of what he wanted from Barker's novels - something that would "introduce and familiarise his Tékumel" readers. Not, say, entertain. Or awe. Or impress. Or any of the other verbs that one could freely bandy about for Lord of the Rings. With that in mind, it it is fair to judge The Man of Gold not as a stand-alone piece of entertainment, but rather as a part of an educational process. The role of The Man of Gold is to sell the reader not on Harsan or his adventures, but Tékumel as a platform for the imagination.
In that respect, The Man of Gold works brilliantly. On one level, a novel - even a clunky one - provides an easy way of delivering dry information in a slightly more palatable way. It is quicker, more accessible, to learn about road systems, dietary habits and economic nuances when they're part of a narrative (no matter how insubstantial that narrative may be). And, on the other, more instinctual, level, The Man of Gold is a challenge: if you don't like the adventure you're reading, you can go make your own. And with a world this compelling, that's a very enticing challenge indeed.
2. Coincidentally enough, there are currently a number of discussions currently taking place around the weighting of 'story-telling' in SF/F, and whether or not there should even be a award specifically focused on this criteria. The origins of the current debate are more about the balance of 'story' and 'politics' - which is already a bit tricksy - are we judging on reader interpretation or authorial intent? But even taking the idea at face value that 'story' is a metric we can single out and measure, there are also cases - The Man of Gold being one of them - where the story sucks for non-'political' reasons.
I'd put my beloved Dragonlance Chronicles in the same bucket: bad story, great series; politics has nothing to do with it. Or, arguably, any book or series, say, the super-popular Stormlight Archives, that combines a glacially-progressing (or simply non-existent) narrative with lavish world-building. I suppose the problem of isolating a single criteria as the basis for anything is that there are always going to be notable exceptions. ↩