A group of friends and I saw Midnight's Children in its New England premiere as part of the Telluride at Dartmouth program at Dartmouth College. (I saw a bunch of the films last year, but don't have time this year and, in any case, am not as enthusiastic about the selection as I was last year.)
The group of us had very different reactions to the movie, with some people extremely enthusiastic about it. For me, it was unfulfilling, and seems a perfect illustration of two general rules: 1.) novelists should not adapt their own books for the screen; 2.) Great books don't make great movies.
A surprising amount of the plot of Salman Rushdie's original novel is retained in the film, and this seemed to me the heart of its problem. A novel of 500+ pages has the room to let its incidents spread out and breathe; a 148-minute film can only include the majority of those incidents if it spends very little time on any of them. And that's what happens. The movie zips along, but it's in such a hurry that nothing much feels like it matters. A story like Midnight's Children, which takes place over many decades and various locations, is especially unsuited to such crammed rushing. It flattens characterizations, making everyone seem like a caricature, and accentuates the many coincidences and contrivances that feel less ridiculous in a large novel. The effect is to make the world of the movie feel absurdly small, and, by the end, to leave it no recourse other than a thin sentimentality. The film sacrifices everything to get as much of the book's plot in as possible, and thus ends up less like an adaptation of the novel than of its SparkNotes summary.
One of my pet peeves with adaptations of complex works of literature is that they rarely seek to find a cinematic equivalent to the literary style. Midnight's Children is interesting as much for its language and structure as for its events, but neither Rushdie in his linear and unimaginative screenplay or Deepa Mehta in her direction find any sort of analogue for that. Rushdie himself gives an ever-present narration in voiceover, further making the film seem like an illustration of the book — we get to look at pictures while somebody reads at us! Last year, I had mixed feelings about We Need to Talk About Kevin, but one of the things I most admired was its determination to be a movie unto itself and not merely an illustration of the novel. The book exists as a work of art in its own right; the movie should, too.
I was often annoyed by the film's colors, which are frequently saturated and sometimes desaturated — manipulation that renders all the whites glowing, blank, and depthless, giving the whole movie an unreal quality that was certainly intentional but to my eyes screamed of kitsch. The same is true for some digital effects in the last third or so of the film, where it sometimes looks like lost outtakes from 300. Occasionally, such as scenes in a Delhi slum, the saturation of colors provides beautiful greens, reds, blues, and yellows, but on the whole the effect was distracting, and especially disappointing given that I thought Mehta's earlier film, Water, was visually powerful and affecting.