The Mill House Murders (1988) - Ayatsuji Yukito

I see the big wheels turnin’
Never endin’, on and on they go

“Big Wheels” (ELO)

This review marks a milestone for the blog, the first full month of regular posts. Granted, that’s only three reviews, but still, it’s something to celebrate…

Though characters being familiar with the tropes of mystery fiction is a common…er, trope of mystery fiction, it’s painfully obvious that most characters who appear in the genre lack this awareness. For example, nobody who was so much as remotely familiar with even the basic trappings of the genre would, after being seriously injured in a deadly car crash, would commission an eccentric architect to build a house in the middle of nowhere and live there in almost complete seclusion, surrounded by a collection of eerie paintings. Yet this is exactly what Fujinuma Kiichi, son of the famous painter Fujinuma Issei, does in Ayatsuju Yukito’s The Mill House Murders. And his dubious decision making doesn’t stop there. Once a year he invites three people to view the paintings, the only times he allows anyone to view his father’s works. In 1985 the annual visit goes awry, ending with one guest dead, another vanished, and a painting stolen. Next year, he invites the same set of people back again. Even ignoring questions of genre savviness, that is clearly not a wise course of action. In addition, another person shows up unannounced, Shimada Kiyoshi, who knew the man who vanished and wants to find out what really happened. The guests go over the events that occurred a year previously, but without giving anything away, it eventually becomes clear just how bad an idea it was to hold one of these gatherings….

The Mill House Murders was published earlier this year, and its release has been eagerly awaited. Mill House is the second entry in Ayatsuji’s Yakata series, the first entry of which, The Decagon House Murders, was published in 2015. Eight years is a long time to wait to continue a series, so there was much excitement when it was announced that Pushkin Press would be publishing the next volume. I haven’t been waiting quite that long to read it, since I read Decagon in 2019, but that doesn’t mean I was any less excited. And in my opinion, Mill House lived up to that excitement, though not perhaps for the reasons you would expect.

Though just directly comparing books with one another is not necessarily the best tact to take in a review, I think that in this case it’s not only warranted, but also possibly intended. Decagon & Mill House are too similar to one another for the correspondence to be accidental. The main similarity lies in the novels’ structure, in both cases two separate but related narratives alternate from chapter to chapter. In Decagon the separation is geographical: most of the cast is on an island getting killed off while a smaller contingent investigates on the mainland. Mill House uses a different kind of separation, one temporal rather than physical. An interesting consequence of this is the way all other variables are held constant. The time is different, but otherwise it’s (mostly) the same people, in the same place, for the same purpose. This allows for another structural point of interest, as for much of the book the action alternates between past events and present discussion about those events. This allows the characters’ theorizing to mirror the reader’s and gives the characters more facts to reason with.

There are also similarities in the plot. A similar (though certainly not identical) type of trick and misdirection are used in both novels. Mill House has far more clues than Decagon did, possibly even too many. Very few readers will reach the summation without unraveling at least a decent portion of the plot. Of course, the very fact that there is a summation represents a structural departure from Decagon. While that novel was absolutely fairly clued, the detective’s reasoning was not explicitly spelled out. Once the reader knew the truth, they would likely remember unnoticed clues, but the chain of reasoning was never explicitly spelled out. That made for a very interesting reading experience, but I can see why it wouldn’t be repeated further in the series. Unfortunately, while the trick in Mill House was very good, the one used in Decagon was a masterpiece. Though the former presents an interesting and satisfying variation on the core idea, the combination of its relative transparency and its similarity to its more accomplished predecessor renders it slightly underwhelming.

Vagrerfgvatyl, gurer vf nyfb na ryrzrag bs fvzvynevgl ertneqvat gur zbgvirf bs gur xvyyref va gur gjb abiryf. Obgu bfgrafvoyl npg bhg bs n fhccbfrq qrfver sbe eriratr, ohg va obgu pnfrf guvf vf whfg n guva irarre bs frys-whfgvsvpngvba pbirevat zber frysvfu checbfrf. Gur xvyyre va Qrpntba pynvzf gb or niratvat gur qrngu bs gur tvey ur ybirq, ohg ur nqzvgf gung fur jbhyq unir orra nccnyyrq ng uvf npgvbaf. Va nqqvgvba, ur xvyyf fbzrbar jub ur nqzvgf unq ab cneg va ure qrngu, fvzcyl orpnhfr vg svgf uvf "cresrpg" zlfgrel abiry-rfdhr fpurzr. Va Zvyy Ubhfr, gur xvyyre pynvzf gb jnag eriratr sbe n pne nppvqrag (juvpu jnf, nf sne nf gur ernqre vf njner, pbzcyrgryl nppvqragny), ohg vg gheaf bhg gung uvf zbgvir fgrzf sebz terrq naq n qrfver sbe gur ivpgvz'f jvsr. Va obgu pnfrf gur xvyyre pynvzf gb or n jebatrq vaqvivqhny, ohg vf va snpg nalguvat ohg. (Spoilers for both Mill House & Decagon hidden by rot13.)

Despite these similarities, there are major differences in the execution of the two books. One which stands out early on is the character writing. In Decagon the characters were fairly flat, with little characterization. Aside from occasional humanizing moments they were mostly there only for the roles they played in the plot. Some of them were downright interchangeable. Honestly, this didn’t bother me at all when I read it, since they were developed to the extent that the plot required, but it’s a criticism that I’ve seen leveled against the book fairly often. Mill House shows a definite improvement in this regard. All of the characters have distinct personalities. They’re not deep, but it marks a definite improvement in Ayatsuji’s style. (This also seems as good a point as any to add that the quality of the translation, done by Ho-Ling Wong, is very good indeed. For reasons that I can’t mention without spoilers, this can’t have been an easy novel to translate into natural sounding English, but it was carried off admirably.)

An unfortunately disappointing point about the novel is that the titular mill house is very much underutilized. The whole premise of the series is that murders keep happening in houses designed by Nakamura Seiji, an architect who always hid secret passages and architectural tricks in his buildings. This allows for any number of interesting situations, as well as allowing completely fair violations of a fundament rule of mystery fiction. As a rule (or rather, a commandment), a writer can’t just have the murder committed via secret passages, as it makes for a remarkably unsatisfying solution. But when the presence of a secret passage is guaranteed, suddenly this is no longer a problem. Not only can it more naturally be integrated into the broader plot, it can also feature in the reader’s theorizing and deductions. In this way something that is ordinarily seen as a cheat becomes an opportunity for new and creative situations and tricks. In this novel, however, the architecture was disappointingly underutilized. Vaqrrq, gur zheqrere arire hfrq gur frpergcnffntr ng nyy, nf ur jnf hanoyr gb svaq vg. And aside from one quite effective set piece, the titular giant mill wheels that line the house also play no role in the plot. To an extent this is understandable, as Ayatsuji was no doubt still deciding what he wanted to do with the series, but it’s still slightly disappointing.

At this point it probably seems as if I'm not quite sold on The Mill House Murders as, although the writing has improved, and although the mystery was certainly good, it just wasn't as gripping or as well executed as Decagon. But that conclusion would be completely wrong, for the simple reason that The Mill House Murders's ending had a real degree of grandeur to it. Mystery fiction, by its very nature, involves taking a situation that seems uncanny and…well, mysterious and explaining it in rational, sensible terms. Some have argued that this makes the solutions inherently disappointing, but I cannot agree with that. A good solution elevates the situation it sprang from, it possesses a satisfaction and a genuine, aesthetic beauty all its own. But it generally (though not always) takes the situation from the level of the wonder-inducing to that of the readily explicable. (Again, this isn’t to say that the solutions don’t inspire wonder, the best ones always do, but it’s wonder of a different sort.) As mentioned, some solutions overcome this difficulty by dint of the sheer brilliance of the solution (as in a great many of Carr’s novels, for example, or in Decagon itself). But others sidestep the issue altogether. In these, the solutions, and the overall situations, signify something in some way, be it artistic, moral, philosophical, or something else altogether. Obviously, I can’t give details, as I would be spoiling some of the best works the genre has to offer, but examples of what I mean include the last chapter of Ten Days’ Wonder, the motive in Gokumontou, and the moment of realization in The Nine Tailors. The solutions to these novels posses a sublimity which is difficult to describe, but which ensures that they stick in the reader’s mind, the relevant scenes indelibly impressed upon his or her memory. And the ending of The Mill House Murders possesses this quality, its very last scene elevating all that has come before. It produces this effect differently than the aforementioned works, as it isn’t part of the solution to the crimes that does so, but the answer to a more…thematic mystery that has come up throughout the book. It reveals an (unclued but foreshadowed) pattern to the events of the novel that gives them grandeur and, within the world of the story, significance. Some might say that the specifics are out of place in the rational world of a mystery novel, but they don’t in any way alter or invalidate the solution, so to me that objection seems to be something of a pointless quibble. While Mill House isn’t on the same level as the books I mentioned earlier, it is of the same kind. And its last line is so good as to almost rival that of Carr’s “The House in Goblin Wood.” Without this closing scene, I could still recommend The Mill House Murders as an entertaining and well-clued mystery that doesn't quite live up to its potential. But with it, I think it’s a minor classic which I recommend strongly. Furthermore, I very much hope that Pushkin will continue with the series, as the next novel, The Labyrinth House Murders, sounds especially interesting.